70-432

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)
1-800-418-6789
SQL Server 2008 Implementation

and Maintenance (70-432)
LearnSmart Exam Manual
Copyright 2011 by PrepLogic, LLC.
Product ID: 012335
Production Date: July 13, 2011
All rights reserved. No part of this document shall be stored in a retrieval system or transmitted by any
means, electronic, mechanical, photocopying, recording, or otherwise, without written permission from
the publisher. No patent liability is assumed with respect to the use of the information contained herein.
Warning and Disclaimer

Every effort has been made to make this document as complete and as accurate as possible, but no
warranty or fitness is implied. The publisher and authors assume no responsibility for errors or omissions.
The information provided is on an as is basis. The authors and the publisher shall have neither liability
nor responsibility to any person or entity with respect to any loss or damages arising from the information
contained in this document.
LearnSmart Cloud Classroom, LearnSmart Video Training, Printables, Lecture Series, Quiz Me Series,
Awdeeo, PrepLogic and other PrepLogic logos are trademarks or registered trademarks of PrepLogic, LLC.
All other trademarks not owned by PrepLogic that appear in the software or on the Web Site (s) are the
property of their respective owners.
Volume, Corporate, and Educational Sales

Favorable discounts are offered on all products when ordered in quantity. For more information, please
contact us directly:
1-800-418-6789
solutions@learnsmartsystems.com
International Contact Information

International: +1 (813) 769-0920
United Kingdom: (0) 20 8816 8036
LearnSmart
Cloud Classroom: Video Training
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Table of Contents
Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
What to Know. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Tips. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Domain 1 Installing and Configuring SQL Server 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Hardware and Software Requirements for Installing Microsoft SQL Server 2008. . . . . . . . . . 7
Installing SQL Server 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
SQL Server Instances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Selecting SQL Server Features. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
SQL Server Security Considerations During Install. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
FILESTREAM Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Data Directories in SQL Server Installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Command Line Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Implementing Database Mail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
SQL Server Configuration Manager. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Domain 2: Maintaining SQL Server Instances. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Understanding System Stored Procedure: sp_configure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Manage SQL Server Agent Jobs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Manage SQL Server Agent Alerts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Manage SQL Server Agent Operators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
The SQL Server Declarative Management Framework (DMF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Backing Up SQL Server 2008. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Domain 3 - Managing SQL Server Security. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
SQL Server Logins and Server Roles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Password Policies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
SQL Server-Level Roles and Database-Level Roles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
The Guest, Public, and DBO Principals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
SQL Server Instance Permissions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Logon Triggers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Impersonation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Cross-Database Ownership Chaining. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Managing Schema Permissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Auditing SQL Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
DDL Triggers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
C2 Audit Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Common Criteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Transparent Data Encryption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
SQL Server Surface Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Domain 4 Maintaining a SQL Server Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Database Management and Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Database Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Database Filegroups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Range of Database Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Recovery Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Detaching and Attaching Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Database Backups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Backup Type Details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Verifying Backups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Restoring Databases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Restore Types and Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Database Snapshots. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Database Integrity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
DBCC CHECKDB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Database Maintenance Plans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Domain 5 Data Management Tasks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Data Export and Import. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Methods of Data Import and Export . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
BCP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
BULK INSERT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
OPENROWSET. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
SQL Server Management Studio GUI Data Import and Export Capability . . . . . . . . . . . . . . . . . 57
Data Partitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Data Compression. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Maintaining Indexes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Domain 6 SQL Server 2008 Monitoring and Troubleshooting. . . . . . . . . . . . . . . . . . . . . . . 62
SQL Server Service Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Concurrency Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
SQL Server Agent Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Finding Error Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Domain 7 Optimizing SQL Server Performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
The Resource Governor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Database Engine Tuning Advisor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Analyzing Trace Data with SQL Server Profiler. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Analyzing Performance Data with Dynamic Management Views (DMV). . . . . . . . . . . . . . . . . . 76
Performance Monitoring With the System Monitor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Domain 8 Implementing High Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Database Mirroring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
SQL Server Clusters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Transaction Log Shipping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
SQL Replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Practice Questions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Answers & Explanations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Abstract
This LearnSmart exam manual will help you prepare for the SQL Server 2008 administration as tested in
Microsoft Exam 70-432. This manual provides practical examples, test tips, and guidance on Microsoft best
practices, which pervade the questions in Exam 70-432.
What to Know
Students should study all facets of installing SQL Server, installing SQL Server instances, and adjusting
SQL installations via SQL Server Management Studio, SQL Powershell, and SQL Command Line. Students
should study SQL query syntax and understand the T-SQL commands related to installation, security, and
management of SQL Server.
Tips
It is important to ensure that you have hands-on experience with SQL Server 2008 and can attempt the
syntax of queries demonstrated in this manual and the configuration examples. The availability of so many
virtual server/PC type products today and Microsofts 180-day free trial of SQL Server 2008 should allow
plenty of chances to work with the product prior to taking the exam.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Domain 1 Installing and

Configuring SQL Server 2008
The first objective when preparing for Microsoft Exam 70-432 is installing Microsoft SQL Server 2008
and obtaining a strong understanding of its various options. As a database administrator knowing how
to install Microsoft SQL Server 2008 accurately in production scenarios will give you the knowledge to
answer questions based on this section of Microsoft Exam 70-432.
Hardware and Software Requirements

for Installing Microsoft SQL Server 2008
Microsoft SQL Server 2008 is available in multiple versions. For test preparation purposes we will focus
on the Enterprise 2008 x64 and Standard x32 versions. Each version of Microsoft SQL Server 2008 has
unique features but for test purposes the installation questions apply to all versions except for Microsoft
SQL Server 2008 Express. This section will detail processor types for Microsoft SQL Server, hardware
requirements, and software requirements. Installation also covers data file locations, default paths for
Microsoft SQL Server, and security during installation.
Microsoft SQL Server 2008 comes in multiple processor versions, and the following definitions are
applicable for understanding the processor versions for installation.
Processor
and System
Architecture Type
Nickname or
Common Reference
x86
32-bit, a 32-bit processor

(e.g., a Pentium IV or
Pentium III processor)
x64
64-bit, a 64-bit processor

(e.g., a AMD Opteron, or Intel
Xeon with EM64T support)
Itanium
IA64, a 64-bit custom

processor built by Intel, which
requires a custom operating
system and versions of SQL
Server. Normally used for highend installations
Example
Operating System
Microsoft SQL Versions

for this Architecture
Microsoft Windows
Enterprise 2003 R2 x86
Microsoft SQL Server 2008

Enterprise 32-bit
Microsoft Windows
2008 R2 Enterprise x86

Standard 32-bit
Microsoft Windows
Server 2003
Enterprise R2 x64

Enterprise (64-bit) x64
Microsoft Windows
7 Ultimate x64

Standard (64-bit) x64
Microsoft Windows
2008 Server Enterprise
R2 Itanium IA 64
Microsoft Windows
2003 R2 64-bit
Itanium Datacenter

Enterprise (64-bit) IA64
Table 1: SQL Server and Supported Architectures

Test Tip: Always ensure a match between the Microsoft SQL Server 2008 version and the server processor
architecture. You are asked to install a 64-bit version of Microsoft SQL Server on a Pentium IV. Is this
configuration possible?
The required SQL Server 2008 Hardware is listed in the table below. The ability to quickly recall these
parameters is an important part of preparing to install SQL Server and serving as a SQL Server Administrator.
The software requirements are addressed and handled by the SQL Server installation wizard.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Required RAM
Required Processors
Required Disk Space
All versions of SQL Server 2008

require a minimum of 512 MB
of RAM. However, SQL Server
2008 will not run very well on
512 MB of RAM.
32-bit versions of SQL Server

require a 32-bit processor with 1
GHz or higher of CPU.
SQL Server has different disk size

options based on the options
selected, but for testing purposes
the amount of disk space
required for the database engine,
replication, and full text search
is 280 MB.
One exception is that SQL

Server 2008 Express only
requires 256 MB of RAM.
2GB of RAM is recommended.
64-bit versions of SQL Server

require a 64-bit processor with
1.4 GHz or higher.
Other SQL Server options will

increase the disk space.
Analysis Services requires 90 MB
of space.
Integration Services requires 120
MB of space.
Reporting Services requires 120
MB of space.
SQL Books Online requires 240
MB of space.
SQL Server manager and other
client components require 850
MB of space.
Table 2: SQL Server Hardware Requirements
Installing SQL Server 2008

Installing Microsoft SQL Server 2008 is accomplished via the setup.exe file of the source binaries.
Installation will run a check for all required hardware and software and stores all events related to the
setup in the following location on the server or workstation hard drive: %programfiles%\Microsoft SQL
Server\100\Setup Bootstrap\Log\. Each file name is named in the format Summary_engine_based_
YYYYMMDD_HHMMss.txt. This is the year, month, date, hour, minute, and second of the setup.
Once installation is complete all of the setup activity is logged in the summary.txt file located in the
above directory. For example, if you are troubleshooting why a certain SQL service or feature is not
available (which the exam will question you on), the first place to start is the summary.txt file.
Test Tip: For any failures during Microsoft SQL Server setup, the summary text file is the place to
start when troubleshooting. This is as easy as opening the file and conducting a search for the
keyword error. If the summary.txt is not listed as an option for troubleshooting the SQL Server
installation, then check for Application Event Log entries, and dont forget the basics for verification
run the SQL Server Management Studio and also check if the binaries are installed.
This overview of a SQL Server installation will cover some of the required concepts for the exam and
illustrate best practices, which Microsoft covers in the exam.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 1: SQL Server Install Options

Figure 1 shows the options for the type of install. We will cover SQL Server failover clustering in Domain
8 in this Exam Manual. The first option will install a new default instance SQL Server, or add a new named
instance of SQL Server. The upgrade option is to upgrade SQL Server 2000 or SQL Server 2005 to SQL
Server 2008.
SQL Server Instances

One very important exam concept is SQL Server instances. SQL Server can be installed on the same
physical hardware with multiple instances that are completely independent of each other. For example,
you can install a default instance of SQL Server 2008 named Server Name. On the same server you can
install a new instance of SQL Server 2008 named Server Name\Test01. This second instance of SQL Server
2008 is completely independent of the first instance for database locations, service accounts, and features.
The only items shared are the core SQL Server binaries installed on the server. The second instance will
browse for applications as a different SQL Server. A .NET application employing connection strings could
connect to the default SQL Server for its core application databases and the SQL Server instance for its
Reporting Server needs. A quick illustration is shown below.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 2: Practical Examples of SQL Server Instances in Use

Test Tip: SQL Server versions can support 8 instances, and SQL Server Enterprise can support 50.
SQL Server instances provide an array of capabilities and are a valid answer to some exam questions.
For example, many of the exam questions focus on upgrading from SQL 2005 to SQL 2008. In some cases
the questions on the exam will have answers that use the least amount of hardware investment.
Anytime that phrase is used you should use an SQL instance to answer the question. For example, if you
are upgrading to SQL 2008 from SQL 2005 and require testing and other certification of existing databases
but only have one server what should you do? You should install a named instance of SQL 2008 on the
same server running SQL 2005. SQL 2005 and SQL 2008 can run on the same hard drives; however, since
SQL 2005 is already running on the server you must use a named instance because a default instance
(the SQL 2005 installation) is already running. SQL 2005 and SQL 2008 use different directories for the
installation so they will not affect each other. This is a rapid way to use an instance to solve an upgrade
issue with constrained hardware and allow detaching of databases running SQL 2005 and reattaching of
the same databases in SQL 2008.
Selecting SQL Server Features

The SQL Server installation allows you to select multiple products. These products are described above in
Table 1 in the disk space section. When installing SQL Server you are faced with this array of functionality
and must understand what each item provides to perform well on the exam.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 3: SQL Server features during install

The SQL Server Database Engine is the heart of the product and provides the location for storing and
processing data securely. SQL Server Replication provides for fault tolerance support by allowing the
mirroring and replication of SQL Server databases to other SQL Servers. Full-Text Search provides
enhancements in the searching of text and content on SQL Servers. Analysis Services is the solution
enabler for analytical processing of data. Reporting Services provides an out of the box reporting solution
for the creation and management of data reports presented via the web. The SQL Server shared features
provide all of the other tools that ensure SQL Server functionality. The important features to understand
for the exam are the management tools, which provide the SQL Server Management Studio.
Test Tip: In normal production scenarios where fault tolerance and high availability is a consideration Microsoft
does not recommend installing all SQL Server options on one server. This is also a performance consideration.
An SQL Server installation requires the selection of services needed for the business, and the selection
of a service will prompt other configuration data as the installation wizard for SQL Server proceeds.
If Reporting Services is selected, a service account and indication of the file locations for the reporting
databases, transaction logs, and binaries will be required.
Test Tip: If users cannot connect to a properly installed SQL Server after installation and all security
is correct and the SQL Server is running, make sure the SQL Server Browser is disabled by default.
By enabling this SQL Service through the SQL Server Configuration manager, clients can connect to the
SQL instance without using port numbers or named pipe names.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
SQL Server Security Considerations During Install

The selection of SQL Server service accounts is required during installation. Service accounts are domain
accounts or local server accounts with required administrative privileges to run the SQL Server services.
Local computer accounts are also available for running the SQL Server services. Local system accounts are
used only by system operations and are not available for an interactive login. The types of local system
accounts available depend on if the SQL Server is a member of an Active Directory domain or a standalone. A stand-alone SQL Server is an installed SQL Server that is in a workgroup versus an Active Directory
domain. System accounts do not require passwords as they are embedded system resources that require
physical or RDP server access to configure. The recommendation (which will help on the test) is to create
service accounts in advance of the SQL Server and make these accounts members of the Administrative
group on the SQL Server. You can then use these accounts to run SQL Server services and ensure you have
an audit trail of which accounts are assigned to what services in an Active Directory domain. An example
service account name is domain name\_svc_dbserver.
Test Tip: Never configure SQL Server services to use the Network Service built-in account in an Active
Directory configuration. This account has rights to traverse domain servers and is considered a violation of
the security best practice.
Figure 4: SQL Server Installation Dialogue for Selection of Service Accounts

Test Tip: The SQL Server Agent is disabled by default. It will not start automatically. We will address this in
SQL Server Configuration later in this section.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Account provisioning is a critical area during SQL Server installation. SQL Server provides two
authentication modes: Windows authentication mode and Mixed Mode authentication, which allows SQL
Server to maintain accounts for SQL Server access. Todays best practice for security is not to configure
any SQL Server in Mixed Mode, thus reducing the attack surface against SQL Server by not allowing
authentication by the SQL Server itself. In some cases a legacy application or support consideration will
require accounts in SQL Server. For the exam, keep in mind that using another instance of SQL Server
to accommodate different security needs is the way to go. Account provisioning is the stage of the
installation when you will decide who will administer the SQL Server. In SQL Server 2008 a decision must
be made to add an administrator to SQL Server for management once the product is installed. This data
should be prepared in advance of the installation. For example, a group in the Active Directory may
contain designated SQL Server administrators or a local group when the Active Directory is not in use.
Test Tip: Make sure you completely understand the ramifications of using Mixed Mode SQL Security! In
most cases dont do it!
Figure 5: SQL Server Account Provisioning and Authentication Mode Choices
FILESTREAM Configuration
An important new feature in SQL Server 2008 is FILESTREAM support. FILESTREAM is a feature that
supports the storage of large images and other BLOB type data over 1 MB. An example is scanned images,
or JPEGs of receipts. This SQL option enhances fast read access for stored objects. This feature can be
configured during installation or by using the SQL Server Configuration Manager. FILESTREAM enables
SQL Server to quickly access objects stored in a server file system versus a database table. The data type
is varbinary (max) when items are stored in file server storage. The configuration requires the name of a
Windows share.
Test Tip: The following T-SQL statement enables the FILESTREAM attribute capability after the installation
of SQL Server. The option 0 disables FILESTREAM, 1 enables FILESTREAM for T-SQL access, and 2 allows
FILESTREAM access for Win32 applications.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 6: T-SQL Statement to complete FILESTREAM post-installation configuration
Data Directories in SQL Server Installation

The location of SQL databases is determined during this phase of installation. Best practices dictate
that Temp DB and the Temp DB log run on LUNs or drive spindles separate from other databases, and in
a RAID 1 configuration. User databases should exist on different LUNs rather than user database log files,
and should also be on different drive spindles. User databases should run on RAID 1+0 volumes, and user
database log files on RAID 1 volumes. Most disk locations should be configured and ready to use
prior to installation.
In common SAN configurations each drive letter will present to the SQL installation as the location to
achieve best practices. For example, the T (T for Transaction Logs!) drive the server has access to is a good
location for transaction logs, configured as a RAID 1 volume. The D drive (D for Data!) is a good location
for data, configured as a RAID 1+0 volume. This allows quick adjustment in the installation to the data
locations and ensures that all databases added will end up configured with the right drive location.
In the high availability section of this manual these considerations control the success of cluster
installations covered in the last section. This section of the installation allows configuration of the backup
directory. This directory becomes important in the setup of management jobs if it is already configured.
Best practice is for the backup location to exist on a protected SAN and to ensure the LUNs are not used in
any SQL read or write operations with the stored databases.
Test Tip: Although a RAID 5 disk configuration does provide fault tolerance, the function of writing
multiple bits across the array is a write inhibiter to SQL Server performance in high volume or high
transaction per second configurations.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 7: SQL Server Installations Data Directories
Command Line Setup

SQL Server allows the adjustment of SQL Server features via the command line after installation of
an instance is complete. The exam has questions related to setup capability at the command line.
The setup.exe file located with the SQL Server source binaries provides the executable to use when
making command line adjustments.
Here is an example of a command line install of SQL Report Server for an SQL instance called BOSS302 in
the MUSTANG domain.
Setup.exe /q/ACTION = Install/FEATURES=AS /INSTANCENAME =MSSQLSERVER/
RSSVCACCOUNT = MUSTANG\RPTSvc /RSSVCPASSWORD = testme01_
This will install SQL Server Reporting Services to the default SQL Server instance, with the assignment of a
Reporting Services startup domain account of RPTSvc and a password.
The best way to prepare for the command line capabilities of an SQL Server installation is to do it handson with scripts. A quick reference for this capability is to open a command prompt, navigate to the SQL
Server 2008 install source, and type setup.exe /?.
Test Tip: The setup.exe command with switches will always use the FEATURES=AS switch. This is
important because on the exam the questions will insert the word FEATURES=IS, which is a quick clue that
the answer is wrong. Also note how the target upgrade SQL Server instance is listed. If it is default, then no
instance specification is required. Read the question very closely and you can identify which answers can
be eliminated easily.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Implementing Database Mail

The SQL Server email feature is important to many database management operations. SQL Server can
email with errors, as a result of SQL application activity, and for SQL Agent job notifications. SQL database
mail is not enabled by default. To configure database email open the SQL Server Management Studio
and expand the Management section. Right-click on database mail and select configure. Clicking the new
profile option will enable database mail for the SQL Server instance in which you are connected.
Assign a profile name and add an SMTP account. The SMTP account data will require information on your
SMTP relay in use or Microsoft Exchange environment. The next option is to configure the profile to be
public, which means it can be accessed by all users of any database, or private which requires access to
use it. Select Send Test E-Mail to test the configuration, and View Database Mail Log if the test is
unsuccessful to troubleshoot.
Figure 8: Send Test E-Mail for SQL Database Mail

Test Tip: The stored procedure sp_send_dbmail operates from within MSDB for system wide database
email operations, so for security purposes ensure each database user that will send database email is a
member of the DatabaseMailUserRole.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
SQL Server Configuration Manager

Once SQL Server is installed the SQL Server Configuration Manager is the tool to use to configure
your SQL instances for service accounts, network ports, and other data such as hiding SQL instances or
disabling SQL protocols. The primary SQL network protocols are TCP/IP and Named Pipes. Best practices
dictate the use of TCP/IP and a selected SQL port. SQL Server defaults to port 1433, but this is easily
changed to support higher levels of security by not allowing SQL to communicate on the default port.
Encryption of connections is a configuration option here and allows the addition of certificates
used for encryption.
Test Tip: Exam questions that require reconfiguration of any network parameters, SQL instance hiding,
and SQL listening protocols should always alert you to look for an answer where the SQL Configuration
Manager is used.
Figure 9: SQL Configuration Manager Showing Protocol Properties for an Instance
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Domain 2: Maintaining SQL Server Instances

Understanding System Stored Procedure: sp_configure
SQL Server provides several tools to manage installations once installed. The system stored procedure
sp_configure, the SQL Server Management Studio, SQL Powershell and the SQL command line.
The exam tends to focus on the use of the T-SQL storage procedure sp_configure. In some instances the
manual will point out when the SQL Management Studio can be used over sp_configure, but this is only
in cases where the change is quicker in SQL Management Studio. The stored procedure sp_configure is a
powerful tool which provides many of the configuration capabilities for adjusting SQL server installation
and activating or changing features. This review will focus on common uses of sp_configure. This is a
system or built in storage procedure that supports SQL tasks across all domains of the exam including
management, security, and performance tuning.
Test Tip: Know what the syntax options are for sp_configure and how to use it to adjust features on
SQL Server instances.
Manage SQL Server Agent Jobs

The heart of managing SQL Server instances is configuring jobs which are managed by the SQL Server
Agent. The exam covers, in detail, numerous fundamentals of using the SQL Server Agent. SQL Server
Agent is the subsystem of SQL Server which runs jobs on a schedule, conducts the appropriate logging of
job activity, fires alerts, conducts database commands, and uses required security contexts for the jobs.
All SQL Server Agent configured jobs are stored in the system database msdb. For exam purposes
remember that the fastest route to backup configured SQL Server Agent Jobs is to backup system
database msdb.
SQL Server Agent jobs are created in SQL Management Studio by right-clicking the SQL Server Agent
jobs tab and following the configuration dialogue flow. The Agent is running and ready for jobs if the green
arrow is present. The job will ask for a name, owner, and a category of database activity (e.g., database
maintenance, data collector, or Full-Text). The steps to the job are inserted next. These steps contain the SQL
commands and syntax needed to complete the required activity. The steps can consist of external code,
operating system support, or Power Shell scripts. Each step allows specification of the database target and
credentials to use for the step. Steps in a SQL Agent Job can access different databases as required and use
different credentials. The Schedule for the jobs is the next configuration step, and it supports complete date and
time selection, recurring, one time, or the option of taking advantage of idle CPU time. An important part of the
selection process is to enable the schedule. If the schedule is not enabled the job will not run. The SQL Server
Agent jobs support alerts that can write to the SQL Server event log or a WMI event. The alert dialogue allows
configuration of severity levels and custom error number. In addition to the alert the job configuration supports
responses with options to the alert, which include notifying SQL operators or running defined SQL jobs.
The Agent will send alerts via email, pager, or a net send command. In most cases net send is outdated.
The Agent will provide notifications of job activity and results via email, page, net send, Windows Application
event log, or provide for deletion of the job. Conditions are set for each notification method, for example,
when the job fails to send an email. Lastly, the target section is configured. The Agent jobs can target the local
server or other defined linked SQL Servers.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 10: SQL Server Agent Job Creation Dialogue

Test Tip: When managing SQL Server Agent jobs, all activity is logged to the file SQLAgent.out file in
the SQL Server log directory, with the full path \program files\Microsoft SQL Server\mssql10mssqlserver\
mssql\log. This file is independent of any alerting or logging setup for the Windows Application event log,
email, or WMI logging.
Scheduling of SQL Agent jobs is a required skill for the exam. SQL Server Agent jobs can have multiple
schedules. For example, the job named 4.6Power 1 and 4.6Power 2 could be scheduled to run the 1st of
the month, 15th of the month, and 30th of the month. When requirements change always determine if the
job schedule must change as well.
Test Tip: Some Agent jobs are dangerous, i.e., they are configured to delete data such as older records or
other data. In some cases the job may be dependent on preparatory activity for tables. As you read the
exam questions, when you spot an SQL Agent job that may delete or remove data and its use is as needed,
make sure you keep the job disabled.
Manage SQL Server Agent Alerts

Once SQL Server Agent jobs are configured the alerts that the jobs create are integral to SQL Server
management activities. Critical jobs for backups, data transformation, data removal, and export of data
(all of which support some business activity) must have alerts to ensure their success or failure is
monitored. In some compliance situations such as PCI Visa certification, this alerting is required and proof
must be available that it exists.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The SQL Server Agent can alert via email, SQL event log, or WMI. WMI or Windows Management
Instrumentation is an alert that is parsed to a system that is running WMI based agents, such as Microsoft
System Center. These alerts can use the rich programming capability of the WMI model. These alerts
can also use any Windows performance object contained in the performance monitoring counter sets.
If you have a case where the alert should monitor SQL Server disk space, then the alert should use the
Windows performance counter element LogicalDisk with a configuration of % Free Space. The solution of
combining SQL Server Agent performance condition alerts with the underlying Windows Server operating
system performance object is very powerful.
Figure 11: SQL Server Agent Alert for Database Disk Size > 80%
SQL Server Agent alerts work in conjunction with other SQL Server notification features. For example,
if you have configured stored procedures running, ensure that they use the option to RAISEERROR WITH
LOG, and this will work with the configured SQL Server Agent alert.
The ubiquitous existence of email enabled cell phones today underscores email notification as a viable
and effective SQLS Server alert method. Emails configured as alerts can notify for both performance
and event alerts. This configuration also builds on the database mail settings covered in the Domain 1
of this manual.
Test Tip: SQL Server Agent logging uses the Windows Application Event log to create SQL Server
Agent alerts. If the Windows Application event log is using the setting to not grow and requires a
manual intervention to clear it, then SQL Server Agent alerts will fail if the Application event log is full.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Manage SQL Server Agent Operators

SQL Server Agent operators are the credentials that receive alerts on jobs. The operator is defined with
the SQL Server Agent job management section. The Operators folder supports the creation of operators
that are available for job notification. The availability of an operator is a key part of the operator options
set and allows for designation of when operators may be on shift, or to support helpdesk operations or
DBA groups. The setup of an operator includes methods to notify the operator, which are via email, pager,
and net send address. Again, net send is a method that will send a message to a logged in computer so
this method is not as reliable as email. The operator view also contains a notification list that stores all
alerts by type sent to the defined operator. The operator objects created for notification are manageable,
thus if someone goes on vacation or is not available that notification operator object can be disabled,
or deleted in the case of an operator change.
Figure 12: SQL Agent Operator Setup

Test Tip: In a scenario where one operator should receive all notifications or the exam question asks
to notify an operator if another operator is not available, make sure you understand the solution is to
configure the fail-safe operator in the SQL Server Agent core properties.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 13: SQL Server Agent Fail-Safe Operator

In cases where database mail is used to provide email services to SQL Server Agent always remember that
if any new database email profiles are created and used for SQL Server Agent notification, the SQL Server
Agent service requires a restart in order to use the new profile. SQL Server Agent restarts do not affect the
uptime of the SQL Server database engine.
The SQL Server Declarative Management Framework (DMF)

SQL Server 2008 includes the ability to create policies that work toward best management practices.
The collection of tools within the DMF focus on the policy capabilities of SQL Server to support ease of
management from one to many SQL Server instances. The DMF targets surface features. For example,
protocol types in use, servers, logins, databases, tables, and users. The policies enforce security and
database settings.
Policy management is achieved via the Policy Management node under the Management area in SQL
Server Management Studio. The policy section supports Microsofts DMF goals for SQL Server, which is to
allow policy control that mirrors the ease and capability of Active Directorys Group Policy features but
is targeted toward SQL Server security, settings, and databases. Policies are created based on need and
the creation of conditions the policy enforces. The conditions are comprised of SQL Server functions and
settings, for example, @RecoveryModel type, and actual calculation functions for determining conditions,
for example, addition, subtraction, and other math functions.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Once a condition is selected in building the expression, the options for that field are available. In this
example, setting a database policy for Recovery Model, when selecting the field @RecoveryModel,
the options of Full, Simple, and Bulk-Logged are available. Conditions can be combined to form multiple
element policies. The policy is applicable against all databases or select databases. Policy configuration
can be restricted based on the formulation of policy facets. A facet can determine exclusions for policies
or other considerations.
Policy Management contains an evaluation feature that allows evaluation of polices prior to production
implementation. Facets of a database are the feature settings, and the SQL Server policies set and manage
facets of a database. Policies can evaluate the conditions they are going to employ and different modes
exist to include:
QQ
On schedule, which is running a policy on a schedule.
QQ
On demand, which is running a policy manually.
QQ
QQ
On change: log, which will log compliance with the policy as it would have done in an
evaluation mode.
On change: prevent, which will stop the policy from executing if a policy evaluation fails.
Figure 14: SQL Server Policy for Database Recovery Model
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Test Tip: Any DMF policy created that is left in the on demand evaluation mode will not enforce the policy.
The policy must be set to on schedule to ensure it is executing. For example, if a policy for auditing
does not allow SQLMail enabling but SQLMail is running the reasons can be that the policy is set to on
demand. When policies execute on schedule, they use the security role of sysadmin, which ensures that
a proper security context is in use to execute the policy. SQL Administrators can log on and evaluate a
policy manually, but this will use the logged on users security context and it may not have all required
permissions to run the policy successfully.
Backing Up SQL Server 2008

The backup and protection of SQL Server data is one of the most important SQL Server administrator tasks.
Companies that lose critical data do not survive for long. SQL Server 2008 provides built in back capability.
The exam has multiple questions concerning the setup, administration, and restore operations for SQL
Server Backup. Any SQL Server database or database log is easily backed up by right-clicking on the
database name and selecting Tasks\Backup. However, this type of ad hoc backup is not reliable nor the
proper method to backup SQL Server production databases. SQL Backups require configuration so that
they are scheduled, automatic, use defined backup sets, and log their results.
SQL Server provides for several types of backups as follows:
QQ
QQ
QQ
QQ
QQ
Full Backup Complete backup of a database file. This type of backup only represents the
database at the time of backup.
Differential Backup A backup that only adds data changed since the previous full or
differential backup.
Transaction Log Full Backup Complete backup of a transaction log. Transaction log backups
can be Full recovery model or bulk-logged recovery model. The full recovery model allows
recovery back to a point in time by restoring backed up transaction logs. The bulk-logged
recovery model is for temporary backups and operations during a large bulk operation.
Partial Backup A backup that only addresses full data in the primary filegroup and any
specified filegroups.
Differential Partial Backup A backup that only contains the data extents modified since the
most recent partial backup.
Establishing a SQL backup is best handled through the DMFs Maintenance Plan Wizard.
This wizard is launched through the Maintenance Plans node under Management in the SQL Server
Management Studio. This wizard allows the setup of multiple schedules when using full and differential
backups, and for different times for a transaction log backup. You can select the Backup Database (Full)
option or others as required. You can also select other common tasks, such as Shrink Database. The wizard
allows configuration of the back location, which can be disk or tape. Normally, backups on disk are then
backed up to tape by another product, such as VERITAS Backup Exec. Compression is also an option to
ensure that data storage is optimized. For full backups the option exists to append to the designated
backup location, or overwrite. In most cases you will append data for the backup so that a point in time is
selectable. The result of the backup job can be emailed or written to a file.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 15: SQL Server Backup Plan

Test Tip: A standard workable SQL Backup schedule is as follows. We are protecting a database called
ponypowersalesDB. This database is backed up with the following schedule and type of backup:
QQ
2300 PM Nightly MTWTFSS: Full Backup
QQ
1000 AM Daily MTWTFSS: Differential Backup
QQ
1400 PM Daily MTWTFSS: Differential Backup
QQ
Every 30 Minutes of Each Day, MTWTFSS: Transaction Log Backup
The exam will challenge you to conduct a database restore for a database problem with
ponypowersalesDB. Pay close attention to the order of the backups and what time the exam indicates the
problem has occurred. In this example, a failure occurs at 3:00 PM or 1500 hours. The restore order is (1)
the full database backup from 2300; (2) the 1400 differential backup; and (3) each transaction log backup
before 3:00 PM. This will restore the database as quickly as possible.
In some recovery cases, the restore will be for a database already restored. The database recovery model
Full Recovery allows you to restore a database using the WITH RECOVERY option. The Full Recovery model
allows the use of transaction log backups, while the Simple recovery does not. This will allow you to
restore a backed up database, which places the database in standby mode and allows the full backup with
recovery option described above.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The COPY_ONLY backup option is also covered on the exam and allows backing up data from a database
or log file without affecting the backup sequence. If you recovered a file within a database from a previous
backup and want to ensure playback of the transaction log backups for true point in time recovery then
restore whatever part of the database file or filegroup is required then backup the transaction log with
the COPY_ONLY option and restore it WITH NORECOVERY and then restore it again WITH RECOVERY. Using
WITH_RECOVERY as the last option in the restore will bring the file back online. This is known as a tail-log
transaction log backup. When the failure of the database in question is reported or known, the quick
backup of the transaction log attempts to recover transactions between the failure and the transaction
log backup. In many cases this will ensure you are the SQL Administrator hero of the day because normally
no data is lost, especially if best practices are followed and the transaction logs are on different disks than
the database files.
Another option, which the exam will also cover, is the SQL 2008 backup option called READ_WRITE_FILEGROUPS.
In database design read-only databases are usually filegroups with the database structure. Backups are
quicker because read-only data is skipped.
Domain 3 - Managing SQL Server Security

SQL Server security questions account for 15% of the exam. This is an area requiring fluency and accuracy.
The exam will expect understanding of SQL Server logins, server roles, the management of users,
SQL Server instance permissions, database permissions, schema permissions, object permissions,
auditing, and data encryption.
SQL Server Logins and Server Roles

SQL Server logins consist of Active Director or domain login accounts, local computer login accounts
when a SQL Server may not be a member of a domain, and the SQL Server account. SQL Server accounts
are logins which are maintained by SQL Server and dont belong to the local computer SAM or an
Active Directory domain. The type of account that SQL Server will allow for a login is determined by
the SQL Server authentication mode. The SQL Server authentication mode is set during install or adjusted
at any time in the Server Properties section of the Server node or with an sp_configure change.
The two authentication modes are Windows Authentication, which will use Active Directory or local server
accounts, and SQL Server and Windows Authentication, which only allows the use of SQL Server accounts.
The latter authentication is commonly referred to as mixed mode. Best practice dictates that only
Windows Authentication mode be used, which reduces server attacks that attempt to compromise SQL
Server based accounts. The default SQL Server security mixed mode login is called sa.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 16: Server Properties for Server Authentication Modes

Logins are created through the Security Node of SQL Server Management Studio. Right-click on the Logins
sub node and select new login. The new login is given a name, but in the case of Windows authentication
this name should correspond to the Active Directory or local server account name. SQL Server Login
creation allows a search of the Active Directory or the local server accounts. No password is assigned for
a Windows authentication login as it will use the password maintained via the Active Directory. A SQL
Server login requires a password and options as to whether to use password policy, password expiration,
and a change on first login. Additionally, SQL Server logins can be certificate based for application support,
mapped to asymmetric keys, or mapped to credentials. A default database for the login is the last portion
of the new login dialogue. Logins can be mapped to the databases they are supporting or should have
access to.
Test Tip: The following syntax will create users in query editor. For a Windows user, add the command
FROM WINDOWS and put the user name in the context of DOMAIN\USERNAME. To force a password
change for a SQL login, add the MUST_CHANGE command to the CREATE LOGIN syntax.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 17: Create Login Command
Password Policies
SQL Server password policies provide the same robust security enforcement as Windows 2003 and
Windows 2008 Active Directory complex passwords enforcement. The parameters include enforcing
password complexity, password expiration, and then enforcing the policy. The policy applies ONLY to SQL
based logins, not Windows Active Directory based accounts.
Test Tip: Use the ALERT LOGIN T-SQL command with CHECK_POLICY to ON to enforce
the password policy.
SQL Server-Level Roles and Database-Level Roles

SQL Server security contains several fixed or default server roles. By placing SQL logins into the available
server fixed roles, the login receives the privileges in which the role is configured. Fixed server roles apply
server wide not just to individual databases. SQL Server also contains multiple predefined database roles
that are applicable to databases individually versus the SQL Server range of a fixed role. An administrator
can create custom database roles, but the predefined ones answer most SQL security issues.
Fixed Roles
Definition
SysAdmin
Any member can perform any action on the SQL Server
ServerAdmin
Any member can set configuration options on the server
SetupAdmin
Any member can manage linked server and SQL Server startup options and tasks
Securityadmin
Any member can manage server security
ProcessAdmin
Any member can manage kill processes running on SQL Server
DBCreator
Any member can create, alter, drop and restore databases
DiskAdmin
Any member can manage SQL Server disk files
BulkAdmin
Any member can run the bulk insert command
Public
Default role for every SQL Server login
Table 3: SQL Server Fixed Security Roles
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
Predefined
Database Roles
Definition
Db_owner
Members have full access
Db_accessadmin
Members can manage Windows groups and SQL logins
Db_datareader
Members can read all data
Db_datawriter
Members can add, delete, or modify data in tables
Db_ddladmin
Members can run any data definition language (ddl) commands
Db_securityadmin
Members can modify role membership and manage permissions
Db_backupoperator
Members can back up the database
Db_denydatawriter
Members cannot add, modify, or delete any data in user tables
Db_denydatareader
Members cannot read any data in the user tables
1-800-418-6789
Table 4: Predefined Security Roles

Test Tip: Exam questions in this area will focus on the least amount of privilege. If the question indicates
that backup capability is required and the ability to kill processes, then assign only those two roles; do not
push up the stack to SysAdmin or ServerAdmin.
In the case of custom database roles, the role should be added to the respective function of predefined
database roles. If you create a database role called Shelby_Backups and want the members to backup
databases, then add the Shelby_Backups role as a member to the db_backupoperator role on the
database in question.
Figure 18: SQL Server Custom Role with Members
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
User mapping is also important to complete the setup of database roles. When setting up a user
administers have the option of mapping the user to a database and selecting a role. Again, best practice is
to use least amount of permissions to maintain a secure audit posture. Logins not mapped to a database
have no access.
Figure 19: User Mapping for a Database Role
The Guest, Public, and DBO Principals

All SQL Server databases include the guest user, which is created anytime a database is created. If a SQL
login does not have an account in a database, then it inherits the guest user rights. By default the guest
user is disabled.
All databases have the database owner or dbo user added upon creation. This role is always a member of
the database-level role called db_owner. This user can configure and conduct all activities in a database.
Each created database is a member of the public database role. Any user who does not have permission
in a database will inherit the public permissions. As with guest user, for security purposes, this user should
remain without rights.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
SQL Server Instance Permissions

The types of SQL Server instance permissions focused on in the exam are logon triggers, cross-database
ownership chaining, and impersonation. The concept of separating security boundaries with SQL Server
instances is in play here. Disparate security measures for password policies, login types, and SQL Server
Authentication mode all should alert you to use a new SQL instance to create good security segregation.
Logon Triggers
Logon triggers are used to activate a stored procedure in SQL Server. Logon triggers are covered on the
exam in the audit and control server portions of the questions. When audit type questions on the exam
describe methods to track user activity the answer is the logon trigger. The trigger can be set to run for
access to the SQL Server or a database. Logon triggers include the ability to activate stored procedures
and ensure they can enforce and prevent actions. A logon trigger can ensure that a login is not running
more than two sessions to an instance at any one time. The logon trigger can enforce (via checks)
conditions about the user session.
Test Tip: Exam questions that ask about restricting user sessions or checking the amount of sessions are
asking about the use of a logon trigger. The logon trigger is created in the master database to ensure SQL
Instance wide enforcement.
Impersonation
SQL Server impersonates credentials that originate as a Windows user when communicating within SQL
Server or SQL services. When SQL runs a process it will use the context of the login, and this transposes to
other functions, processes, Linked SQL Servers, and database access.
SQL Server has a Common Language Runtime (CLR) built in to execute managed code within the SQL
Server memory space. The SQL.Context.WindowsIdentity in the CLR is used to impersonate a Windows
account with permission to access the required databases and code. This code call will set the context
under which the code will execute which allows creation of Windows accounts with the appropriate
permissions to execute in the CLR space.
Test Tip: Exam questions that offer answers for impersonation by adding the Windows account to the
Active Directory Domain Admins group or local server Administrators group are wrong answers.
The enforcement of security always dictates the least amount of privilege to execute successfully.
Always evaluate exam questions in the security arena with a focus on least permissions.
Cross-Database Ownership Chaining

Database application solutions sometimes require the access of multiple databases in a SQL
Server instance. In cases where this is required a chain of ownership is created across the required databases.
The second database in the chain will not evaluate permissions of an object that was previously
authenticated in the first member of a chain. This does provide for a speed increase in some cases;
however, this feature is disabled by default.
Test Tip: Cross-Database chaining is enabled server wide via the SQL Server Management Studio and
Server Properties, or with sp_configure cross db ownership chaining =1 to ON. The recommended solution
for exam and management purposes is to use ALTER DATABASE to turn on cross-database chaining or
individual databases that require it rather than setting it server and instance wide.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Managing Schema Permissions

Another key securable in the SQL Server security ecosystem are database schemas. Database roles,
such as dbo or other logins, own or manage schemas throughout the database realm. A key enhancement
with SQL Server 2005 was the introduction of user and schema separation. SQL 2008 maintains this
security split. Schema objects include tables, stored procedures, views, and roles that can access and
manage these interactions. The practical effect of this security feature is that a SQL login named
Shelby302 can own a schema named SchemaPower, which contains a table, a stored procedure, and a
user-defined function. The SQL login has access to and ownership of all of these database objects through
singular schema ownership.
Test Tip: In SQL 2008, it is simple to change the owner of a schema from one login to another. This is the
least disruptive administrative path. If a user named Shelby302 leaves the company, a simple change of
ownership to the new login completes assignment of all the objects in the schema.
The inclusion of the schema object creates the four-part contents of a fully qualified object name in
SQL Server 2008; for example, BOSS302.HEADERS.SCHEMA1.TABLE1. In other words, the SQL Server
name, database name, schema name, and the object. SQL Server 2008 supports the use of synonyms to
access these four-part defined objects. For example, the code below creates a synonym of BOSS for the
BOSS302.HEADERS.SCHEMA1.TABLE1, and then conducts a select of all data.
Figure 20: Creation of a Schema Object Synonym

Test Tip: The ALTER AUTHORIZATION versus the ALTER USER command quickly changes
schema ownership. The movement of schema objects is accomplished quickest through ALTER SCHEMA.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Auditing SQL Server

The auditing of SQL Server is one the most important day-to-day tasks of an SQL Administrator, and this
is covered in multiple converging areas on the exam. A collection of methods is reviewed below, which,
taken together, present the picture of a valid SQL Server audit posture. From PCI (Payments Card Industry)
certification to HIPPI (Health Care Regulations) the storage of data, which is comprised of personal or
financial information, dictates that SQL administrators must have a command of SQL audit methods.
SQL Server can audit activity and send the output to multiple locations to include files, the Windows
Application Log, and the Windows Security log.
Test Tip: SQL Server can only access the local servers Windows Security log if the SQL Server service runs
as a Local System, Local Service, Network Service, or an Active Directory based account which contains the
rights to access the log.
SQL Server allows creation of objects for auditing via the CREATE SERVER AUDIT command. An audit
condition is set up with this command and a destination as described above is selected. Once audit
conditions are created, the SQL administrator can add events to the condition with CREATE SERVER AUDIT
SPECIFICATION and CREATE DATABASE AUDIT SPECIFICATION commands. If as an SQL administrator you
were tracking failed logins, you would create an audit specification that tracks the failed login and adds it
to the audit file and location.
The following are examples of standard SQL Server audit specifications and can be applied to all users:
QQ
SUCCESSFUL_LOGIN_GROUP
QQ
LOGOUT_GROUP
QQ
FAILED_LOGIN_GROUP
QQ
AUDIT_CHANGE_GROUP
QQ
SERVER_PERMISSION_CHANGE_GROUP
The following are examples of database specific audit specifications and should be applied
to all databases:
QQ
DATABASE_CHANGE_GROUP
QQ
DATABASE_PERMISSION_CHANGE_GROUP
QQ
DATABASE_OBJECT_ACCESS_GROUP
The following are database table specific audit specifications that ensure auditing is applied at the lowest
point of the database object:
QQ
SELECT (this audit feature captures the SELECT statement not the data!)
QQ
INSERT
QQ
UPDATE
QQ
DELETE
Protection is also required for the SQL Server audit files. Ensure that you are familiar with auditing the
reading and changing of the produced SQL Server audit files. An audit set on master.sys.fn_get_audit_file
records any login that reads the audit file or attempts to change it.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Each database has a property called Change Tracking, which is enabled at the database object level and
then available to any table within the database. This will track all changes to data in a database within a
table and allow you to specify an interval to retain the change data.
Figure 21: Change Data Tracking Configuration on Database Table
DDL Triggers
DDL triggers are similar to logon triggers in that they execute events based on some defined
audit condition. DDL triggers provide the power to enforce SQL Server auditing by ensuring code
execution for defined events and the execution of alerting when required. DDL triggers track specific
events within databases. For example, if an SQL administrator wanted an event written every time a
database object is deleted a DDL trigger would satisfy this requirement.
DDL triggers help collect audit activity in databases. For example, an SQL administrator always has the
option of creating an audit table based on an XML column and inserting any DDL trigger events into the
table for simple parsing and reporting later.
A very powerful feature of DDL triggers is the ability to act on CREATE_USER actions and then execute
procedures to enforce permissions management. For example, if you wanted to have all users for a
database receive data_writer capability on creation the DDL trigger can import the user name from the
XML based data and then form an SQL statement to execute the requirement.
Test Tip: When the exam question asks how to audit an event every time it occurs and write an event,
the answer should focus on the DDL trigger capability.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
C2 Audit Configuration
The C2 audit mode was a previous SQL Server audit posture that matched event logging for both failed
and successful attempts to access statements and objects. This mode is easily turned on in SQL Server
Management Studio\Security Properties or by using sp_configure. All C2 information is stored in the sys.
traces catalog view.
Test Tip: C2 auditing collects a lot of data. If the file location where the logging is saved runs out of
disk space then SQL Server will shut down. To bring SQL Server back up you must free up disk space,
or for the exam know that you can bring up SQL Server and bypass logging with the following command:
Run: sqlservr f.
Note that the C2 audit standard is now replaced in SQL Server (and the industry) by Common Criteria.
Common Criteria
Common Criteria is an industry security standard certification that is promulgated by the NSA DBMS
Protection Profile. SQL Server 2008 is compliant when using Common Criteria compliance as an
option to EAL1+.
Common Criteria addresses the following elements. Residual Information Protection (RIP) ensures the
overwriting of bits in memory with a known pattern of bits before the memory is used by another process.
This prevents buffer and other types of data harvesting from leftover processes in memory. Common
Criteria supports the viewing of all login statistics to include successful logins, unsuccessful logins, number
of login attempts, and login times. These statistics are viewed in the sys.dm_exec_sessions Dynamic
Management View (DMV). The last element of Common Criteria is that any table with DENY will take
precedence over any column object of that table with a GRANT.
Common Criteria audit options are enabled via the SQL Server Management Studio or with sp_configure.
You must download a script from the Microsoft SQL Server Common Criteria web site to complete the
configuration and restart the SQL Service.
Test Tip: When exam questions focus on logging all login attempts both successful and unsuccessful,
the answer will include enabling common criteria compliance. When you review the answers to this type
of question take note that logon triggers only execute AFTER logons have occurred and only capture
successful login activity.
Transparent Data Encryption

Securing SQL Server data with encryption helps further the security model. SQL Server 2008 provides
for Transparent Data Encryption or TDE, which provides the required cryptographic functions to encrypt
and decrypt data as it is read from or written to SQL storage. This is a huge advantage because it does not
require that the application use SQL Server for encrypting and decrypting data. TDE uses a DEK (Database
Encryption Key), and this key is stored in the database boot record in case recovery is needed.
Test Tip: Anytime a database using TDE is moved to another SQL instance, the database encryption key
requires a backup and restore to the new SQL instance. If you do not also get the certificate, you cannot
restore without the key because TDE database backups are also encrypted.
A range of encryption capabilities exist with Windows 2008 and Windows 2003 which have BitLocker
Drive Encryption (BDE) and the Encrypting File System (EFS), respectively. This defense layer is only at the
operating system level and not within the SQL Server application. SQL Server databases do not achieve
encrypted operations by only encrypting the disks with the operating systems. TDE is the answer that
provides encryption, and it is preferred because it does not affect applications, such as ASP.NET web
applications using a SQL database.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The use of TDE requires underlying key configuration for SQL Server. You will need to create a master key.
This is accomplished via the CREATE MASTER KEY statement. The next requirement is a certificate, which
like all certificate operations, can be generated from the Active Directory Certificate Authority, the local
server certificate authority, or a third party certificate from VeriSign or a similar accredited certificate issuer.
The ability to create the required DEK for TDE is contingent upon the above two steps.
Figure 22: Enabling TDE for a SQL Database
SQL Server Surface Area

The SQL Server surface area is reduced by default to ensure that services that are not necessary after
installation are not running. The SQL Server surface area is the portion of the SQL Server services
exposed for use and connections via tcp ports or named pipes. An example service is database mirroring,
which uses multiple ports to function so it is disabled by default and reduces the SQL Server surface
area. The configuration of security for the surface is area is addressed by sp_configure and the previously
mentioned SQL Server management policies. A number of configuration facets exist which allow
definition of policies relating to the SQL Server surface area. The facet Surface Area Configuration contains
the key surface area services for enabling or disabling. The creation of a condition using the facet allows
creation of a policy to enable or disable it.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 23: SQL Server policy condition using the Surface Area Configuration facet
disabling Service Broker
Once a condition is created based on a preconfigured facet this condition rolls up to a policy on
the server, for example, a combined security policy which will include the defined conditions, in this case,
those related to surface area configuration.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 24: Adding the configured condition to a policy for Surface Area Configuration
Test Tip: In this area, the exam focuses on reusing policies for other SQL Server instances. SQL Server
Management Studio allows the export of a facet as a policy to a file. On a new SQL Server instance
select the exported policy file, evaluate it, and apply for any violations. Sp_configure is the answer
when exam questions ask you to evaluate a surface area solution for one SQL Server instance versus
multiple instances.
Domain 4 Maintaining a SQL Server Database

Database Management and Configuration
Questions on the exam from Domain 4 cover a wide range of everyday SQL Server administrative tasks.
These areas include backups, restores, database file types, checking database integrity, recovery models,
and ranges of database options.
Database Files
The fundamental building blocks of SQL Server are the collection of files that make up the complete
database object. Anytime a database is created two files are created by default. A database consists of
a primary database file, with the extension *.mdf, and filegroups within the database object, with the
extension *.ndf, and the log file, which has the extension *.ldf. A database object always has a primary file,
and then filegroups become secondary. Within each SQL database object are the various elements:
tables, indexes, and schemas.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 25: Database Files with extensions, Primary, Filegroups, and Logs
SQL database objects can contain multiple log files. When one log file fills up, the next log file is used.
This ensures that the database object stays online and does not suffer downtime.
SQL Server database object files should be stored on separate disks or arrays to achieve fast throughput.
The primary *.mdf and *.ndf files should be on different disks or arrays than the *.ldf files. This supports the
goal of disk read and write operations employing different I/O paths to different physical disks. Ideally each
disk subsystem path will use a different disk controller or Host Bus Adapter (HBA). The SQL Server database
filegroup of *.ndf files can also benefit from performance with storage to different disk subsystems. In the exam
questions note that RAID 1 is the best disk RAID type to support a log file. This also provides fault tolerance.
Performance experts do not agree about the use of large RAID 5 volumes as locations for SQL Server *.mdf
and *.ndf files, but this does provide a solid fault tolerance solution. In most cases high transaction based SQL
Server and OLAP implementations use RAID 10 (RAID 1 + 0) SAN configurations to achieve the fault tolerance
capabilities of RAID 5, with the speed of RAID 1 data reads and writes.
SQL Server 2008 stores data on a page. This is a common term used in backup and restore scenarios
because at times the restore problem is to just restore a corrupted page holding some data within the row
on that page. Standard page storage is 8 KB. The grouping of 8 pages together in an SQL Server database
file structure forms an extent, which is 64 KB. The exam expects an understanding of SQL Server database
structures to this level and questions emerge on the smallest elements of a row and table.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Database Filegroups
SQL Server database filegroups build on the common database file structures listed above and expand
the options for addressing data segregation, performance, and recoverability.
A filegroup stores multiple SQL Server database files in it. The creation of a database establishes the first
file in the group as the PRIMARY. SQL Server administrators in the design process can create multiple
filegroup objects which become the *.ndf files listed in the definition of database files. Filegroups allow
the spreading of tables and indexes across physical files. Exam questions from the SQL Server Performance
section (Domain 7) have filegroups as the basis for achieving great optimization in very large database
implementations. You can create multiple filegroups and place them on different volumes and achieve
better I/O. When queries execute to various tables spread between filegroups performance is increased.
Filegroups also provide the foundation for good database maintenance and backup planning.
Read-only data is placed into a read-only filegroup, which reduces backup time. One strategy is to place
large data in different filegroups and then use full backup or differential backup at different times to
protect the data. For example, a filegroup named PARTS is backed up fully on Monday and Thursday,
and a filegroup named SALES is backed up fully on Tuesday and Friday. This strategy saves time because
only portions of the database are backed up and on different days to avoid congestion when backing up
large amounts of data.
Test Tip: For exam purposes note that a transaction log can never be in a filegroup!
Figure 26: Database with multiple files in a filegroup
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Range of Database Options

Each database created in SQL Server has a full set of options available once created. Options can
be changed on databases with the SQL Server Management Studio or with the ALTER DATABASE
T-SQL statement. The exam will expect complete familiarity with the SQL Server Management Studio
database options section and with using ALTER DATABASE.
The ALTER DATABASE command has the following database options available:
LearnSmart
QQ
Optionspec
QQ
Auto_option
QQ
Change_tracking_option
QQ
Cursor_option
QQ
Database_mirroring_option
QQ
Date_correlation_optimization_option
QQ
Db_encryption_option
QQ
Db_state_option
QQ
Db_update_option
QQ
Db_user_access_option
QQ
External_access_option
QQ
Parameterization_option
QQ
Recovery_option
QQ
Service_broker_option
QQ
Snapshot_option
QQ
Sql_option
QQ
Termination
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 27: ALTER DATBASE command in action (changing recovery model to full)
Recovery Models
The type of recovery model used in the SQL Server database configuration will drive the restore method
available for SQL server databases. SQL Server has three recovery models: full, simple, and bulk-logged.
Each of these models has a different effect on how the databases transaction logs are set up. All of the
recovery models are set in the Database Properties\Options node. Right-click on a database in SQL Server
Management Studio and select database properties.
The simple recovery model dictates that the transaction logs are not kept. When database data is
changed and committed the transaction log in simple recovery mode is also cleared, or in SQL Server
terms, truncated. When a truncation of the transaction log occurs it is no longer useful in the restore
of a database. When a database is in simple recovery mode the only database backup type available is
full backup. An SQL Server administrator in support of production databases would not use this recovery
model. If a database is backed up each hour over six hours and a failure occurs at 5 hours and 55 minutes,
the data loss window is now 55 minutes of data as the administrator would restore the backup taken
at hour 5.
The full recovery model does not automatically create SQL checkpoints on the transaction log and
truncate them. The full recovery model allows the introduction of two new backup and restore functions:
differential and transaction logs. In the full recovery model an administrator can restore one file,
without the entire database, or one or more pages and a transaction log. The full recovery model enables
full point in time recovery for a database for any time range covered by the backed up transaction logs.
The bulk-logged recovery model is like the full recovery model except that this mode supports bulk
operation and is designed to support the full recovery model. This model is a performance-enhancing
mode for support of intensive database operations such as bulk imports of data or index creation. Bulk log
will reduce the log space required. However, one key difference from the full recovery model is that the
bulk-logged model does not support point in time recovery!
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Exam Prep: The 302Info database in Simple Recovery Mode

The 302Info database is configured with the simple recovery database option. Each night the 302Info
database is backed up at 0200. One day you come to work and the Sales department reports that their
.NET applications are reporting the error Database 302Info Not Found. You check the SQL Server instance
and discover a failed drive. The drive failure occurred over 10 hours after the last backup, and the company
had conducted about 10,000 sales between the last backup and the drive failure. You are new to the job
and not held responsible for this catastrophe. However, the Sales Executive Vice President tells you:
QQ
Never allow the loss of more than 10 minutes of sales data.
QQ
The company does not have much available storage space for backups.
What solution will you provide?

Solution: The 302Info database in Simple Recovery Mode
The first part of the solution is immediately apparent: the 302Info database must switch to the full
recovery model. This is accomplished easily through SQL Server Management Studio and database
options for 302Info. This change will not disrupt current database operations and is seamless to users.
The second big requirement is to not lose more than 10 minutes of sales data. This requirement suggests
scheduling a transaction log backup every 10 minutes.
The last requirement is that the recovery method cannot use much disk space. Transaction log backups
at such a short interval (10 minutes) will use a minimal amount of space since they only capture what has
changed in that time span.
This is another graduated walkthrough of a question type that may appear on the exam.
Keep the following in mind when answering these types of questions:
QQ
QQ
Some of the answers will offer the differential backup as an option. This option, however,
uses much disk space and backs up everything changed.
Answers will also offer different time intervals. Always ensure you match your answer with the
solution. In this case the operative timeframe is 10 minutes.
Detaching and Attaching Databases

The SQL Server Management Studio supports the easy detaching and attaching of SQL databases.
This function is primarily used to move a database to a new server. A database detach will ensure
that no processes or applications have the database in use and ensure that the files can be easily
copied and move. Once a database is detached navigate to the file system directory containing the
databasename.mdf (database file) and the databasenamelog.ldf and copy them to the new SQL
Server drive location.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 28: Detaching a SQL Server Database
Database Backups
Database backups and restores occupy a large portion of an SQL administrators time during support
or production database operations. This area is also crucial to successfully passing MS Exam 70-432.
Domain 4 adds to the basics of backups and restores covered in Domain 2. In this section we will
investigate more scenarios for backup and restore questions versus core definitions and how to set up
and conduct these services.
Backup Type Details

When you are ready to take the MS 70-432 Exam, you should already have a solid understanding of
backup and restore definitions.
Exam Prep: Backup and Restore for 302Info
You are responsible for a database called 302Info. The database is approximately 460 GB, and data exists
in multiple filegroups. This full backup takes 6 hours. Only 30% of the data in the 302Info database
changes daily. The Parts filegroup is read-only and rarely updates.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The business division, which uses the 302Info database for corporate sales, has requested that you devise
a backup implementation that achieves the following business goals:
QQ
QQ
QQ
No more than 2 hours of data can be lost at any time.

The time window to back up the database, which is currently at 6 hours, is too long and
must be reduced.
If any database recovery is required the business team demands that it be as quick as possible.
Filegroups for db.302Info
Filegroup Types
PowerAccessoriesSales
Primary (150 GB)
Parts
Read-Only (200 GB)
CustomerInfo
Read-Write (110 GB)
Solution: Backup and Restore for 302Info

The scenario above will have four possible answers on the exam, but in evaluating this question,
several backup concepts emerge quickly as clues to the correct answer.
The use of a read-only filegroup called Parts immediately signals that the backup option READ_WRITE_
FILEGROUPS is okay to use for a partial backup, i.e., the backup command will only backup read-write
filegroups and the primary filegroup. By using the T-SQL statement BACKUP DATBASE 302Info READ_
WRITE_FILEGROUPS TO CompanyBackupSAN, you achieve a backup of the changed data each day,
but now avoid backing up over 200 GB of static data. This will reduce the backup time to less than 3 hours,
which is a 100% reduction in time. When data changes in the Parts filegroup, then schedule a backup to
accommodate the new data and then move forward again with the READ_WRITE_FILEGROUPS option.
This is a partial backup type.
The next portion of the question focuses on how much time can exist between data failures. So if
a failure of the database occurs at 1410, and it is backed up every two hours with a differential option,
then recovery can occur within the expected two-hour time window. The solution is to set up a BACKUP
WITH DIFFERENTIAL every two hours for the 302Info database. This will ensure that no more than two
hours of data is at risk throughout the day. For a restore, the fact that the full backup did not back up the
200 GB read-only filegroup along with the smaller size of the differential backups ensure that you will
enjoy a faster recovery time.
The combination of the above two solutions allows you to quickly isolate the answer on the exam.
Make sure you focus on answers with solutions that do not work so that you can quickly identify the
correct answer. Examples of solutions that dont work for the above scenario are as follows:
QQ
QQ
QQ
BACKUP DATABASE 302Info to CompanyBackupSAN. This solution maintains the current

situation of a long backup timeframe.
BACKUP DATABASE 302Info to CompanyBackupSAN WITH DIFFERENTIAL, and do this nightly.
This solution fails because it does not create the base backup required for recovery and as the
partial backup base.
Backing up the databases every four hours without using a differential, which supports a quick
restore, or not backing up the databases within the requested data loss prevention interval,
which in this scenario is two hours.
Test Tip: The above scenario demonstrates a method you can use to evaluate the requirements of a
question on the exam and how to eliminate wrong answers and isolate the correct one.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Verifying Backups
Verifying backups in SQL Server is an important administrator task and one that should happen often to
ensure that when disaster strikes you are ready. The exam will expect you to know how to verify a backup
during the backup process and also manually.
The Verify backup option is available when setting up a backup job through the SQL Server Maintenance
Plan interface. This option will ensure that the backup is complete and the volumes are readable.
Another verification method is the option of performing checksum before writing to media.
The backup process will check media but continue on when it finds an error, but not use that portion
of the backup volume.
Figure 29: SQL Server Backup options for reliability and verification
Restoring Databases
The restore process for SQL server databases takes several routes. Keep in mind for the exam that the
SQL Server database recovery model dictates a route for the restore operation. If the recovery model
is set to simple, then a transaction log restore is not possible because it was not an option in the
backup configuration.
SQL administrators will excel on the exam by understanding the following restore methods, and this will
translate to successfully supporting corporate SQL database operations. The first time you conduct a
restore from well managed backup options that save a company money and time you will be the hero!
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Restore Types and Strategies

The following are restore strategies with an explanation of how the recovery model of the database
influences them:
QQ
QQ
QQ
QQ
Restore of the entire database.

Simple recovery model: Restores the entire database with no transaction log capability.

Full recovery model: Restores the entire database, and allows restore of any
subsequent log files to the full backup. As with all full recovery model restores the last
restore of a log is restored with the RESTORE WITH RECOVERY option.
Restore of a data file, which can include different filegroup database files and read-only files.

Simple recovery model: Restore individual files only if the database has at least one
read-only filegroup.

Full recovery model: Restore individual files, without restoring the entire database.
Filegroups being restored are always offline.
Restore of a Page, which can correct a damaged page or pages.

Simple recovery model: Not an option.

Full recovery model: Page restoration relies on having an unbroken chain of log
backups to the current log file.
Piecemeal restores recover databases in stages at the filegroup level, starting with the primary
and working through the remaining read/write or secondary filegroups.

Simple recovery model: supported.

Full recovery model: supported.
Exam Prep: Database 302info Restore

The 302info database is 400 GB total and has 300 GB of database files located on SAN drive Y and 100 GB
of log files located on SAN drive L. You have implemented the following backup schedule:
Backup Type
Backup Time
Full Backup
2300 Nightly
Differential Backup
1000 Daily
Differential Backup
1400 Daily
Transaction Log Backup
Every 30 Minutes, Daily
At 1310 PM on a Tuesday, the SAN array hosting the database file abends (abnormal end to operations)
and is offline. How will you quickly recovery the database?
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Solution: Restore Scenario

On the exam quickly identify your options for the restore. In this case you must note the complete loss of
the SAN drive supporting the database file. The SAN drive supporting the log file is still up and working,
however, this is a scenario that requires a restore to get the database back.
Keep the following steps in mind when performing a restore:
1.
2.
3.
4.
First and foremost back up the transaction log!

Restore the last full backup.
Restore the latest differential backup if applicable.
Restore each log file backup taken after the last differential backup in sequence, with the
RECOVERY option selected, until restore of the last transaction log.
When thinking about the solution above, using these steps you can successfully restore the database back
to its state at 1300 or a data loss window of 10 minutes.
The backup of the transaction log is detailed below in the Transaction Log Tail Restore section. This last
transaction log backup prior to the restore allows the recovery process to even restore some of the
transactions between 1300 and the time of the failure at 1310. There is a differential backup conducted
at 1000 daily so this will reduce the amount of time for the restore due to not having to restore logs from
2300 the night before.
Rule out answers that do not back up the log as the first step of the restore process. Beware of answers
that suggest that a restore of the latest transaction log is the only requirement for logs. Also note
answers that restore transactions logs from the original full backup time and do not incorporate the
differential backup.
Database Snapshots
Database snapshots provide another level of recoverability and fault tolerance in protecting SQL Server.
Microsofts database snapshot technology builds on their expertise with virtual drive snapshots for
Windows operating system products, but adds the important SQL unique snapshot requirements of
ensuring that the database snapshot is readable by SQL Server. A snapshot is a point in time picture of the
state of a database and all of its objects. Database snapshots are extremely useful for code updates that
affect table structures or other database stored procedures. When a code update goes awry or some new
process removes a table or index by mistake, the reversion back to the previous database snapshot saves
time from a restore and ensures rapid recovery capability.
Test Tip: There is no capability within the SQL Server Management Studio graphical tools to create
a database snapshot.
The database snapshot is read-only. However the contents of the created snapshot are usable with T-SQL
commands for restore back into the source database.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Snapshots are created with the CREATE DATABASE command. An example is shown below.
Figure 30: CREATE DATABASE command for a database snapshot

The above snapshot command created a copy of the database Superchargers. When any data changes
in the database Superchargers, that data will show up in the created snapshot. The snapshot has a utility
for restoring corrupted data or any of the objects in the database that are corrupt. Database snapshots
receive data from the primary database via copy-on-write, which is how the snapshot stays up to date.
Test Tip: Database snapshots will grow in time just like the database in which they are taken. Database
snapshots must always exist on the same SQL instance as the source database.
Database Integrity
Database integrity refers to the usability of SQL Server databases. Are they fit? Are they corrupted?
The primary tool in your SQL administrative tool kit for analyzing database integrity is the
DBCC command. Common SQL Server challenges for databases include torn pages that affect
performance and data retrieval. Abrupt power outages can contribute to database damage.
The DBCC command and its options allow a range of checks and repair activity.
DBCC CHECKDB
Database Console Commands or DBCC are used to conduct maintenance and integrity checks
on databases. DBCC has a set of statements related to validation, which help ensure database integrity.
The primary statement used for validation is DBCC CHECKDB. This check will ensure correct allocation and
structural integrity of all objects in a database. It will identify and repair a wide range of possible errors.
Exam Prep: Using DBCC CHECKDB
You are responsible for a database named 302Info. This database is on an SQL Server 2008 instance.
302Info has six filegroups, and tables located over several filegroups. A problem has occurred and a
developer has indicated a check of the databases integrity is needed as quickly as possible.
What will you do?
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Solution: Using DBCC CHECKDB

This question is like many posed in the exam in that some of the data given in the question is
not relevant. The DBCC CHECKDB command has existed since SQL Server 2000, so the version of SQL
Server is irrelevant. The filegroup locations in the question are also irrelevant since they wont affect the
DBCC CHECKDB command.
The quickest answer to this question is to run DBCC CHECKDB 302Info. This will provide a quick check to
verify the integrity of the database.
As you work through the exam questions some will be just this easy. Dont second-guess yourself
when determining the answers as this type of question can lead to overthinking. DBCC CHECKDB
has many options. You can use the WITH PHYSICAL_ONLY switch, which will only check and not
repair any problems.
An exam answer for this solution using the REPAIR_FAST option is a quick clue not to use it. This option
only applies to other versions of SQL Server, not SQL Server 2008.
Database Maintenance Plans

This domain focuses extensively on database files, options, backups, restores, and methods to ensure
databases are available for use. The Database Maintenance Plan capability of SQL Server 2008 with
both a manual and wizard driven creation process is a feature that SQL administrators will use often.
Maintenance plans can set up backups of all types, index reorganization tasks, database compression,
and perform other required daily tasks.
Figure 31: Preconfigured Maintenance Tasks
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Maintenance Plans are built from the Maintenance Plans node in SQL Server Management Studio.
The wizard supports rapid point and click creation of maintenance plans with all required options,
which include time, schedule, databases to affect, types of backups, and notifications via text files and
email reports. The large number of predefined maintenance tasks facilitates the quick build of multi-step
plans to allow individual step schedules.
The Maintenance Plan designer allows the manual creation of database maintenance plans and editing
of created plans.
Figure 32: SQL Server Maintenance Plan Designer

Test Tip: When exam questions pose the task of conducting multiple actions with the
least administrative effort, this points to using a SQL Server Maintenance Plan. Only members
of the sysadmin role can create and modify maintenance plans.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Domain 5 Data Management Tasks

The manipulation of data within SQL Server is what the database product is all about. The storage and
retrieval of data is paramount to supporting applications and other information systems. The exam
focuses heavily on these skills and as an SQL administrator so will your day-to-day activities.
Data Export and Import

SQL Server provides multiple methods to export and import data. We will review Bulk Copy
Process (BPC), BULK INSERT, OPENROWNSET, and the extensive set of tools available within the SQL Server
Management Studio. SQL has a dedicated serviceSQL Server Integration Services (SSIS)designed to
address the export and import of data with GUI-based support and SQL T-SQL statements. Many of the
scenarios on the exam and on the job include activities that require movement of data from SQL Server
2005 (and possibly SQL Server 2000) instances to new SQL Server 2008 instances. Data import and export
can be part of established business processes or a single task. Data movement also focuses on hydrating
SQL Server Reporting Services databases, data warehouses, and other analytical operations.
The first consideration for exporting and importing data is preparing the data. Whether using the BCP
utility, BULK INSERT, or OPENROWSET, the data requires proper formatting for either import or export.
The following conditions apply for data preparation for export and import:
Export
QQ
QQ
QQ
Never use a name for a BCP export if the data file name is already in use. This will OVERWRITE the
current exported data in the file.
The SELECT permission is a minimum security requirement to export data.
When exporting data, SQL Server 2008 will use parallel scans for export retrieval, i.e., data is not
retrieved in any specific order. When the exam mentions specific order for an export, always use
the queryout option explained below, with an ORDER BY switch.
Import
QQ
The file must be in row and column configuration.
QQ
Character data and native binary format is supported, including Unicode.
QQ
The destination table for the data import must already exist.
QQ
The target table must have a corresponding column with the imported data.
QQ
Use XML formats to import fixed length fields with a format file.
QQ
LearnSmart
CSV files are not supported for BULK IMPORT operations, however, if data fields do not contain
a field ender (the comma normally in a CSV file), or none of the data is in quotation marks,
then CSV-based data may import.
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Methods of Data Import and Export

One of the most common data import and export activities is copying data between servers. The BCP
utility, BULK INSERT T-SQL commands, and OPENROWSET all play a role.
BCP
The BCP command can export or import data from a SQL Database object. The BCP utility is command line
based, and accessed via bcp.exe. This utility is beneficial in that it runs on a small memory space and is fast.
When exam questions focus on quick out of band SQL Server data import or export--think BCP! It can be
accessed via the Windows command line, in script, or with SQL Server Powershell.
Figure 33: BCP command line with switches

The BCP utility requires an exact order of commands and is case sensitive. The BCP utility also requires
a format file to set options for the data import or export.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The most commonly used syntaxes for BCP are listed below:
BCP Setting
Explanation
-a packetsize
Number of bytes in each packet for import/export, default is 4096 bytes.
-b batchsize
Number of rows to copy, default is ALL.
-c
Specifies the use of character mode for data transfers. This supports a non-SQL
destination and will output data in ASCII text.
-C codepage
Sets the relevant codepage to use in the transfer (does not matter when character
values are > 127 or < 32).
-e errfile
File to store error messages.
-E
Uses identify values.
-F firstrow
Determines the number to use for first row.
-f formatfile
A very important component of any BCP command syntax. The format file is a file
with an extension *.fmt and contains key information for the import or export job.
-h loadhints
Hints consist of TRIGGERS during the job, ROWS_PER_BATCH, or CHECK_

CONSTAINTS, and ORDER.
-l inputfile
Used in interactive mode for responses to command prompt questions.
-k
Keeps nulls.
-L lastrow
Will set the last row.
-m maxerrors
Determines the number of errors which can occur before stopping the BCP.
-N
Determines native export for non-character data and Unicode.
-n
SQL native mode, used in transferring data between SQL Server

instances or databases.
-o outfile
File for BCP output during scripted BCP use.
-P password
Password for login during BCP.
-r rowterminator
Determines when to terminate a row.
-S servername
The SQL Server name or instance.
-T
Allows use of a trusted connection.
-U username
Login name during BCP.
-w
Sets wide character use (Unicode) for imports and exports.
-x
Supports the creation of an XML file instead of a standard file as listed in

the f switch.
Table 5: BCP Settings
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
A sample export command with BCP using a trusted connection is listed below. The data exported is
headers information from a table called Parts within the 302info database. The T switch ensures a trusted
connection at the command line.
bcp 302info.parts.headers out headers.dat T
Exam Prep: Using BCP
You manage several databases, one of which is called Accounts. This database has several tables in which
customers require the data monthly, and the data is very large. When exporting this data using SQL Server
based SSIS packages, you notice it takes a very long time and affects the capacity of your SQL Server
during the export. Since your operations are 24/7, any latency inducing operations are not good. What
actions should you take?
Solution: Using BCP
The answers on the exam will focus on many methods for exporting and importing and appear to always
mix in an XML file to help control data that is not formatted correctly. Other web service terms such as
XSLT and XSD do not APPLY to data import and export with BCP. The focus must be on time and efficiency,
which BCP adheres to since it runs in separate memory space, has the capability of running queries for
data, and is fast. As you evaluate the possible answers remember that BCP is always good for offsite
movement of data because its modes can support character-based export since the target system may
not be a SQL Server. Never forget that BULK INSERT is as advertised; it can only import data. BCP is very
scriptable in Cscript, Powershell, and SQL Agent jobs.
BULK INSERT
The BULK INSERT command is very similar in syntax to BCP but only addresses import, not export of data.
Note for the exam that BULK INSERT capability is related to setting the BULK-LOGGED recovery model.
Use this recovery model during BULK INSERT to speed up the process with less transaction log writes.
Database recovery models are flexible and can shift between BULK-LOGGED and FULL with no issues.
A sample BULK INSERT is listed below where table data is inserted into the Parts table object of the
database headers from a file named lineitem.tbl, with a field terminator switch and row terminator switch.
Figure 34: BULK INSERT Command Example
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The advantages of BULK INSERT are that it is great for importing large amounts of data, and the T-SQL
syntax switches are very similar to BCP. BULK INSERT has a corresponding recovery model to enhance
the insertion of very large amounts of data: BULK-LOGGED. This command also works very well with
constraints that include NULL values, clustered indexes, and foreign key constraints.
OPENROWSET
OPENROWSET provides another SQL Server 2008 method for data import. OPENROWSET is a function that
is used among SQL Servers for one-time, ad hoc data access. It is an OLE DB function called with the FROM
statement, or INSERT, UPDATE, and DELETE T-SQL commands. OPENROWSET also can read bulk data from a
file and then return it as a row set. This command allows rapid return of a row from a source and insertion
of the data into the destination.
The following example demonstrates the use of OPENROWSET with a SELECT statement and SQL Server
OLE DB as the provider. Different data providers may be invoked as required. The provider is listed first,
then the connection information, followed by the SELECT syntax, the FROM syntax, and additional
commands as required. One implementation note for using OPENROWSET is that Ad Hoc Distributed
Queries must be configured to on, or a state of 1. This is disabled by default.
Figure 35: Example of OPENROWSET Data Import
Figure 36: Enabling Ad Hoc Distributed Queries
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Exam Prep: Using OPENROWSET

You manage a database named 302info on a SQL Server 2008 instance. The sales team requires a data
import from another server running SQL Server 2005 on an as needed basis. You set up an OPENROWSET
function to complete the task but each day it fails. How should you troubleshoot this?
Solution: Using OPENROWSET
This type of exam question requires a quick examination of the answers for why OPENROWSET is failing.
The first area to check in SQL Server 2008 is the state of Ad Hoc Distributed queries, which is turned
off by default. Make sure you enable this. The answers will take on multiple paths. One path is almost
always the use of SQL Server Integration Services (SSIS), but in this limited ad hoc case, the solution
is an OPENROWSET one. SSIS provides a much more extensive bulk data import and export
transformation capability.
SQL Server Management Studio GUI Data Import and Export Capability
SQL Server 2008 provides a rich point and click wizard to set up data import and export packages via
tasks established in the SQL Server Management Studio. The graphical import and export wizard creates
packages that can import and export data in an ad hoc as required manner or as a scheduled job.
The SQL Server import and export wizard created packages can be managed by the SQL Server Business
Intelligence Development Studio and are SSIS packages. This wizard is accessed by right-clicking a
database in SQL Server Management Studio and selecting Import or Export under Tasks.
Figure 37: SQL Server GUI Data Import and Export Wizard
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The point and click wizard allows setting all package elements for the data import or export activity.
Key settings created in the wizard include:
QQ
Data Source (SQL Server, OLE DB, Microsoft Excel, or Microsoft Access)
QQ
Authentication Methods
QQ
Destination Server
QQ
Data copy options to include one or more tables or views, or a query to establish content
QQ
Table Mappings
QQ
Preview of Results
QQ
Destination Options to include deletion of rows, or appending
QQ
The ability to run the package immediately or save it for future use as an SSIS package
QQ
Ability to send exported data to multiple formats to include flat files, Microsoft Excel, and
Microsoft Access formats
Test Tip: Exam questions that describe regularly scheduled table data copies and import or export of data
from Microsoft Excel and Microsoft Access to SQL Server are indicating the use of SSIS.
Data Partitions
SQL Server 2008 tables and indexes can exist in more than one partition. This is termed horizontal
data partitioning because groups of rows are organized into individual partitions based on an
indentifed column. This type of table organization is critical to large database efficiency and performance.
Partitioning considerations include if the table will host data used in different ways, and any query or
other database operations that are latent or not executing properly. Additionally, you can partition a table
using the concept of vertical partitions, which is spreading the columns of a table across multiple tables
versus the rows.
If a large table for auto parts exists called PARTS, and the primary database operations against it are
INSERT or DELETE (some of the data is only used for SELECT queries), a partition could occur on the PARTS
types, and isolate the items used in SELECT queries from the ones used in INSERT and DELETE functions.
Another good example would be an order table, in which the current months data was used primarily
for INSERT or DELETE operations. The previous months data, which is mostly used for SELECT queries,
represents a target for partitioned tables. This ensures database functions do not scan or query an entire
large table.
The following commands support the partition function: CREATE PARTITION FUNCTION, ALTER PARTITION
FUNCTION, and DROP PARTITION FUNCTION. The exam focuses on an understanding of these commands
and when to use them to manage a data partition. Each new partition must have a range called the split
range, and the values must be unique from the existing table. If a partition function takes in the year 2006
for orders, then the new range must start with a different range, 2007, for example. The merging of data
into the new partition is accomplished once a range is created.
The steps for creating a table partition are:
1.
2.
3.
LearnSmart
Create or use an existing partition function that sets the correct range boundaries
(CREATE PARTITION FUNCTION).
Create or use an existing partition scheme tied to the partition function.
Create the table using the partition scheme.
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
This sample function demonstrates the creation of four partitions for each year with an int column.
CREATE PARTITION FUNCTION myDateRangeFP1 (int)
AS RANGE LEFT FOR VALUES (2006, 2007, and 2008)
This sample function demonstrates the alteration of an existing partitioned table in which orders received
in 2009 would split out.
ALTER PARTITION FUNCTION myDateRangeFP1 ()
SPLIT RANGE (2009)
A data partition is easily dropped with the following command:
DROP PARTITION FUNCTION myDateRangeFP1
Exam Prep: Altering a Data Partition
You manage a large database named PARTS hosted on SQL Server 2008. The PARTS database has a table
called INVENTORY, which is partitioned on its ReceivedDate column using the PFReceivedDate function.
This partition uses one month values.
You need to write a script that moves the older data for more than one year to the partitioned table for
PFRecivedDate into an empty partition. Which code should you use?
Exam Solution: Alerting a Data Partition
This example alerts you to use the ALTER PARTITION FUNCTION with appropriate syntax and switches.
You will run ALTER PARTITION FUNCTION PFReceivedDate() SPLIT RANGE(datelastyear). This will split out
the data from last year. Then you can merge the data into the empty partition, and add to it as needed.
Test Tip: The partitioned table feature is only available in SQL Server 2008 Enterprise and
Developer versions, not Standard.
Data Compression
SQL Server 2008 has several methods for data compression within the database object itself. The exam
expects an understanding of when to employ data compression based on the situation and how it will
affect the parameters of the database object.
SQL Server 2008 row compression allows the storage of fixed length data in variable length format,
and page compression attacks data redundancy.
SQL Server 2008, unlike SQL Server 2005, extends the capability of row compression to all fixed length
data types, to include integer, char, and float data types. This type of support allows data compression
without the application using the data requiring any type of change to access it. The goal is to reduce byte
storage consumption. If a common integer can fit in 1 byte, there is no reason to use 4 bytes to store it.
By using variable length storage format 3 bytes are saved. Variable characters are now a standard option
on column definitions.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 38: Example so variable character data types

Page compression is the first place to start when ensuring data within a SQL Server database is
compressed and uses less space. Page compression first acts on a row, and then dictionary based
compression for multiple instances of the same values. When databases have duplicate data in many
of the fields, space is conserved with this method of compression. In most exam questions on data
compression clustered indexes are in use because in a large table they would exist anyway. When page
compression is used on a table heap, any existing pages are not compressed, only new ones. A rebuild of
the heap is required to ensure that the compression is used for all pages. On the exam when instances
of clustered indexes are in use for tables with compression ensure that any maintenance conducted to
support compression drops the clustered index first.
Exam Prep: Data Compression
You run a database called Headers. The Headers table includes products that are dependent on other
products for use. You need to store the number of dependent products for each product. For example:
1 to 10 Products.
The 302Info .NET web application requires the ability to do an efficient sort based on the number of parts.
You need to create a NumberofParts column. What is the solution?
Exam Solution: Data Compression
This solution is focused on using varchar capability within SQL Server 2008. Due to the requirement
of accessing filtered data, the solution must include use of a filtered index on the created column to
provide for the number of dependent parts. Filtered indexes on sparse columns support efficient sorting
on the data. You should create the NumberofParts column as a sparse column of type varchar, which will
enforce data compression and reduce storage.
For questions like this on example on the exam look for answers that use Null in their solution as
this would immediately discount them. Also watch out for answers using non-clustered indexes in
the solution. In this example a non-clustered combined with nullable columns would result in no
storage savings. Also note you cant add a sparse column to a clustered index.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals
1-800-418-6789
Maintaining Indexes
SQL Server administrators must understand all types of indexing for database tables to ensure accuracy
when answering exam questions. The types of SQL Server indexes include:
QQ
Clustered Indexes
QQ
XML Indexes
QQ
Non-clustered Indexes
QQ
Composite Indexes
QQ
Unique Indexes
Certain index types are linked to the column type in use. A spatial index can only be used with a spatial
data type. A spatial index as with other indexes requires a primary key prior to creation. A spatial index
supports unique SQL Server data type features in the area of geometry and geography, which use the
spatial data types. Indexes are an underlying performance enhancement, which support rapid queries
and speed of data access. SQL Server database tables can contain clustered or non-clustered indexes.
A clustered index will keep the actual data rows at the leaf level of the index. Normally, they are in
ascending or descending order. Only one clustered index can exist on any one table or view. The term
heap refers to a table that has no clustered index think slow or small! Non-clustered indexes differ from
clustered indexes in that they dont actually contain the data in the rows, but only pointers to the data.
Non-clustered indexes cant be sorted, but you can create multiple non-clustered indexes per table
or view. In SQL Server 2008, 999 non-clustered indexes are supported. The composite index and the
unique index are two more forms of SQL Server table indexing. A composite index will contain multiple
columns, up to 16. A unique index, as you may have guessed, is based on a unique table parameter such as
a primary key. For example, when you create a primary key in a table an automatic unique clustered index
is created. XML columns in tables require an XML index which will create an indexed view of all the XML
tags, values, and paths of the XML data.
As an SQL administrator you will need to understand the commands available in SQL Server 2008 to change,
delete, or add indexes. These tasks are always included in other SQL Server maintenance operations. They are
also included in performance tuning as dropping and rebuilding indexes is a fairly common SQL Server
administrator task. The exam will expect fluency with the following T-SQL commands:
QQ
CREATE INDEX
QQ
CREATE SPATIAL INDEX
QQ
CREATE XML INDEX
QQ
DROP INDEX
Exam Prep: Alter Index

You support a database called 302EnginesDB. It contains a table called PonyCars. The following table
describes the index setup of the table. IX_PRIMARY is a clustered index. The other listed indexes are
all non-clustered.
LearnSmart
Column Name
Type of Data
Data In Indexes
CarID
Char(10)
IX_PRIMARY
IX_CARTYPE
IX_CARPRICE
Type
Varchar(25)
IX_CARTYPE
Description
XML
None
Value
Decimal
IX_CARPRICE
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Performance testing indicates you need to add the Description column to the IX_CARTYPE index as an
included column. Table downtime should be minimized. What is the solution?
Exam Solution: Alter Index
You would use the CREATE INDEX WITH DROP EXISTING and perform this command offline.
The requirement to rebuild this index offline is due to adding an XML column, which is a large data type.
You can rebuild the index online if a large data type is not included. Anytime a column is added to an
index you must drop the existing index. In this case ALTER INDEX will not work because you cant add
a column with this command. In exam questions that focus on solving slow query time due to index
fragmentation, ALTER INDEX is a solution with the REBUILD WITH ONLINE switch.
Test Tip: Exam questions that focus on index management will always use DBCC commands in
possible answers. Take note that these commands, while good for database maintenance, do not
address most index management issues.
Domain 6 SQL Server 2008

Monitoring and Troubleshooting
Troubleshooting SQL Server problems is primarily an SQL administrators job and will factor into the
exam questions. The ability to rapidly solve SQL Server issues is related to exam objectives in this domain
and being able to rapidly troubleshoot problems. The areas of SQL Server service statuses, problems
created by concurrency such as a dreaded deadlock, SQL Server Job Agent issues, and how to quickly find
error information all contribute to problems that occur with SQL Server 2008 and how to quickly assess
and remediate those problems. In production situations SQL Server 2008 downtime is not an option, and
this domain will demonstrate methods to quickly solve problems that affect uptime.
SQL Server Service Problems

SQL Servers uptime is related to the functioning of the product features that are installed.
These features are listed as services and reflect the range of SQL Server capabilities. The most important
service is the SQL Server database engine, which is termed the SQL Server (name of instance) service.
A quick look in the SQL Server Configuration Manager under the SQL Server Services shows the status of
all SQL Server Services. Services include the SQL Server Browser, SQL Server Agent (name of instance),
SQL Full-text Filter Daemon Launcher (name of instance), and any other installed SQL services,
for example, SQL Reporting Services (SRS). SQL Server status depends on logons, the start mode defined,
and that no system issues or SQL Server configuration has stopped SQL Server from starting.
At times the problem is that the SQL Server service will not start. You need to immediately troubleshoot
this, and the quickest way is by changing the sqlserver.exe file to use the f switch to start up. The f switch
starts SQL Server in a minimal configuration mode and allows a single connection to the SQL Server.
This type of startup mode can address multiple issues, for example, when logging has filled up the drive
it is assigned to and the SQL Server service cant start up. This option is easily accessed through the
properties section\advanced tab of the SQL Server service in the SQL Server Configuration Manager.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 39: SQL Server Advanced Startup Parameters

If the SQL Server Agent is set to start automatically, this must be disabled prior to launching the SQL
Server service with the f option. This is due to the f mode option only allowing one connection to the
SQL server. If the SQL Server Agent starts automatically this service will use the one allowed connection to
the SQL Server service.
SQL Server provides the option of a Dedicated Administrator Connection (DAC) to make a connection
to SQL Server 2008 when normal connection attempts fail. DAC access only works on the server itself.
However, you have the option to use the sp_configure stored procedure to configure the DAC to allow
remote connections. But this must be enabled prior to a problem that requires this special connection.
The DAC is usable via sqlcmd and requires the special administrator switch A. The best strategy is to use
the DAC to connect to the master database, which is usually available when problems occur and the SQL
instance is running.
Troubleshooting SQL Server services also requires an understanding of service startup types.
The SQL Server Browser service allows the use of the SQL Server Resolution Protocol (SSRP) to resolve
which port SQL Service is listening on. This obviates the requirement to use standard SQL Server ports
such as 1434, and allows other port use to avoid scans for SQL Server services by malicious programs.
Additionally, this helps support multi-instance SQL Server use for the same physical service. The SQL
Server Browser service is not set to start by default. If you plan on using this, then set the properties of
this service in the SQL Server Configuration Manager to the automatic startup type. The exam will pose
various questions that focus on service use such as the SQL Server Browser service, which doesnt
start automatically. In this case there is no need to over think the question, just focus on ensuring
required services are started automatically.
Test Tip: Services like the SQL Server Browser service do not require extensive permissions to run.
When evaluating exam questions make sure you find solutions that are paired using the SQL Server
Browser service with the Local Service account versus the Local System or the Network System account.
The Local Service account has a reduced permission set compared to the Local System account with
extensive permissions, which the SQL Server Browser service does not require.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
C2 audit mode and the common criteria along with other logging choices can impact disk space and
serve as the cause for SQL Server failure. When logging this much data, disk space is impacted. If smaller
drives are used as the location for audit log files and the physical disk fills up, SQL Server will not start.
The solution on the exam in this case can range from starting SQL Server with the f switch to moving the
audit files to another drive with more space.
The account under which an SQL Server service runs also presents an area for troubleshooting. If you
use all Active Directory domain accounts as the service accounts for SQL Server services and any Active
Directory group policy options require the Active Directory accounts to change the password (or they
expire), this will result in a failure of the SQL Server service using that Active Directory account.
Normally, you can use Powershell scripts or other management scripts to change SQL Server Active
Directory account passwords within the required time, which ensures that the account is changing
passwords per company policy or best practice. However, when an Active Directory account expires,
locks out, or requires a password change the SQL Server service in question can switch to Local System
to get SQL Server started again immediately while troubleshooting. Again, for exam purposes,
remember that the SQL Server Configuration Manager supports rapid change of SQL Server service
account information.
Exam Prep: Troubleshooting SQL Server Services
You have two Microsoft SQL Servers on your network. One is an SQL Server 2005 instance, and the other
is an SQL Server 2008 instance. SQL Server is configured with default connection information. The SQL
Server Agent runs on both SQL instances.
A new corporate application will access both SQL Servers for its data. You are working with the application
developers to troubleshoot any connectivity problems that may arise.
The developers report that they can only find connections to the SQL Server 2005 instance, but are forced
to specify a port when connecting to the SQL Server 2008 instance.
You are required to ensure both SQL Servers can return connection information automatically without
specifying a port. The solution should not change the established network security. What is the solution?
Exam Solution: Troubleshooting SQL Server Services
In this type of exam question the challenge of connecting to SQL Server 2008 is presented. The service
in question should immediately come to mind--the SQL Server Browser service. It is not configured
to run automatically in SQL Server 2008, which explains why the developers cannot enumerate or find
it automatically. The simple solution is the correct one--you need to start the SQL Server Browser service
and configure the service to start automatically. Answers that focus on changing an Active Directory
service account are not correct. Other answers on the exam for this scenario tend to focus on the SQL
Server Agent service, which again is not related to this issue.
Concurrency Problems
SQL Server operations sometimes suffer latency or dropped transactions due to blocks, deadlocks, or locks.
These are problems that require rapid identification with the SQL Server 2008 tools available.
As SQL Server reads and writes data to a database locking occurs to support concurrent operations. This is
SQL Servers method to ensure that the same data is not written to the same row/page at the same time.
This is also the heart of shared database access by users and applications.
The exam focuses on the tools available to troubleshoot concurrency problems and when to use them.
The SQL Server tools for diagnosing concurrency problems are the SQL Server Profiler, SQL Dynamic
Management Views (DMV), and the activity monitor.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Blocking occurs when a lock is created and then escalated based on the size of the scan. For example, a
scan of data in a table that exceeds 5,000 locks will cause a lock escalation and ultimately block the next
process requesting the reading or writing of data. Blocks by default do not timeout. You can monitor SQL
Server for block conditions using the SQL Profiler and filter on the event (Lock: Escalation event),
the Dynamic Management View (DMV), or regular polling of system information for lock conditions.
These tools will help gather the information on blocks such as the block and lock time per object,
which will lead to a plan to remediate the condition. The SQL Profiler is available within the SQL Server
Management Studio and allows you to record information over time for a deeper analysis of the
blocking issue. The results are recorded to a trace file and are available for examination.
Test Tip: On the exam, questions focusing on time as a component of troubleshooting the lock issue
suggest the use of SQL Profiler.
Figure 40: SQL Server Profile Lock Events for Tracking Block Conditions
Deadlocks occur when several tasks within the same SQL Server memory space block each other by
having a lock on a resource that another database function is trying to lock. This situation is caused by
a variety of resources to include locks, worker threads, memory, parallel query execution, and multiple
active result sets. SQL Server is configured to automatically search for deadlocks every 5 seconds with
the database engine deadlock detection scheme. The resolution to a deadlock is made by the database
engine and is a result of choosing a thread to continue. This results in an error 1205, which is prevalent in
the exam questions on how to track deadlock conditions. Additionally, the trace flags 1204 and 1222 exist
to document deadlock conditions that have occurred. You can turn on trace flag 1204 as a standard event
to write to the SQL Server error log, which will allow parsing of this error by your management activities
and response to the occurrence. The flags list all relevant information for the deadlock to include what
process was running, the SPID, the mode, the row in question, the object, the key, and other metadata
that is available. When using SQL Profiler to track deadlock activity ensure you select the Deadlock Graph
as an option.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Exam Prep: Troubleshooting SQL Server Deadlocks

You manage a SQL Server 2008 instance. Customers report that purchases often timeout
with a strange error. You must quickly determine which resource is causing this problem.
How can you determine the problem quickly?
Exam Solution: Troubleshooting SQL Server Deadlocks
In this type of exam question you must monitor the SQL Server processes and determine what process is
causing the issue. This is conducted by the SQL Server Profiler and uses the deadlock graph event. This will
provide detailed historical data on the problem and allow you to view the process causing the deadlock.
Many exam answers in this area will suggest the use of a dynamic management view with tran lock.
However, this only shows current locks and does not reveal any history. For example, you may need to run
the SQL Profiler and create a trace file that spans the entire time period in which the problems appear.
The Activity Monitor also seems like a good choice for this type of problem, but it does not provide
historical data either, only a just in time view of what is locked currently. Sometimes the answers will also
have an option to use sp_lock. But note that sp_lock is a retired legacy command that will not provide any
help with troubleshooting a deadlock issue.
Dynamic management views, in general, provide views and functions to return server state information
to support health monitoring, diagnosing of problems, and performance tuning. The sys.dm_tran_locks
dynamic management view is very important when troubleshooting calls for quick views of transaction
locks at various times of the day. Think of this dynamic view as a snapshot for quick views and the ability
to rapidly gleam information about a deadlock condition. Figure 41 below demonstrates the use of the
above dynamic management view to display lock information for a database with an ID of 5.
Figure 41: Sys.dm_trans_locks Dynamic Management View
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
SQL Server Agent Problems

The SQL Server Agent is responsible for running many of the configured management jobs for SQL
Server and associated jobs to support database applications, data movement, and a host of important
database tasks. When the SQL Server Agent fails, rapid troubleshooting must ensue to correct the
problem, especially when the SQL Server Agent defined job is critical to production operations. As covered
in other domains, the notification features of the SQL Server Agent are powerful and ensure that you
know when a failure occurs.
One of the main areas for SQL Server Agent job failures is the security context in which the job is run,
or security contexts for steps in a job. SQL Server Agent can use a proxy account to run jobs, which are the
credentials of a Windows user. The SQL Server Agent can use the security context of the established proxy
user to impersonate it and run the job. SQL Server Agent job failures can fail quickly based on the role of
the user configured for the job and SQL Server role of the user, or the roles of the user within msdb.
SQL Server Agent proxy accounts are created in the SQL Server Management Studio\SQL Server Agent
object, under the Proxies section. Select the subsystem for the proxy account upon creation.
Figure 42: SQL Server Agent Job Proxy Account Creation
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The EXECUTE AS command is also a security context to support SQL Server Agent job execution. In some
cases the proxy account or other security context may not have the correct permissions to execute a
T-SQL job step. The EXECUTE AS command allows the use of a security context to complete the T-SQL step.
The SQL Server Agent proxy cant be used with a T-SQL job step because it is a Windows user account.
The stored procedure sp_enum_proxy_for_subsystem provides a return of which proxies have
permission for which SQL Server Agent job subsystems. This is another useful tool in troubleshooting
permission problems.
Another area of focus for the successful execution of SQL Server Agent jobs is the ability to connect to
required SQL instances. In multi-instance SQL Server configurations the SQL Server Agent can serve in a
master role, and then target other SQL Server instances for job execution. However, to ensure that jobs
work without SQL Server port or instance connectivity issues make sure you create a SQL Server alias for
the target servers. Additionally, during the creation of the Master/Target SQL Server job setup you can
specify the servers for connections.
Helpful diagnostic information for the status of steps in SQL Server Agent jobs can be obtained by
configuring each job step to log detailed information to a table. The contents of this table can be accessed
quickly with the following stored procedure: sp_help_jobsteplog. The SQL Server Agent core options
support the settings for the maximum number of rows entered for each job and the maximum size of the
log about the job.
Finding Error Information

SQL Server 2008 provides several areas for rapid access to error information. These areas include: the error
log, the agent log, the job execution history, and event logs. SQL Server administrators will study these
logs daily to quickly diagnose problems, and the exam will test your knowledge of where to find relevant
information for reported problems.
The SQL Server error log is the first place to check for problems with SQL Server. This log file is located
in the default location of the first SQL Instance, which is \\installdrive\program file\Microsoft sql server\
mssql10.mssqlserver\mssql\log\errorlog. The most recent file doesnt have an extension. This log will
contain SQL Server specific errors, not the type of errors you may find in one of the Windows Server
event logs. The size and location of the error file is adjustable via the SQL Server Configuration Manager.
The SQL Server Management Studio has a log file viewer utility under the Management node,
which supports the aggregated view of several critical SQL Server logs. They are the Database Mail log,
the SQL Server logs, the SQL Server Agent logs, and the Windows NT logs of the server hosting
SQL Server. The log file view contains a filter setting to rapidly search the logs for keywords, such as
the user, computer, and text contained in the message and a Source. Additionally, any of these logs are
exportable for further review or upload into other systems.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals
1-800-418-6789
Figure 43: SQL Server 2008 Combined Log View with Filter Settings
Custom error information can be written to the SQL Server logs or Windows event logs. The management
stored procedure xp_logevent writes user defined messages to a log destination. Stored procedures in
use in SQL Server can harness this capability to ensure database errors are meaningful and assist in the
troubleshooting process. Anytime a stored procedure raises an error message, they are automatically
stored in the sys.messages table. A query against this table provides any errors logged by stored
procedures regardless of their severity level.
Exam Prep: Finding SQL Error Information for Job Execution
You manage 10 instances of Microsoft SQL Server 2008. You have one instance that has a problem every
hour with the SQL Server Agent service. Troubleshooting this problem requires detailed information about
the job execution. Execution trace messages should write to the SQL Server Agent error log. How will you
achieve this with in the quickest manner?
Exam Solution: Finding SQL Error Information for Job Execution
This exam question requires SQL Server Management Studio. A quick right-click on the SQL Server Agent
node will make the change to this tool easily. This type of question will present solutions that focus
on triggers that do not address this logging question. Other solutions suggest using DBCC as a form
of turning trace flags on. Again, this does not address configuring the SQL Server Agent to write trace
execution messages.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 44: Setting the Execution Trace Messages Option for SQL Server Agent
Domain 7 Optimizing SQL Server Performance

SQL Server 2008 performance is always important in real world use and for this exam. SQL Server
responses that are latent or fail impact the applications and solutions that use the database tier.
Imagine ecommerce solutions that fail to return product inventories or process orders as consumers
are clicking and waiting. SQL administrators are expected to rapidly identify problems and correct them
through data analysis and trending of SQL Server operations. The tools covered in this domain provide,
in depth, the data that classifies performance issues and fixes them quickly.
The Resource Governor

The SQL Server Resource Governor is a member of the collection of SQL Server 2008 policy initiatives.
The resource governor allows the definition of groups or resource pools that use SQL Server resources,
and the placement of classifiers or policy statements on the group. A resource pool is comprised of
the physical resources of the server--CPU and memory.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The processing of constraining a resource works as follows:

1.
2.
3.
4.
5.
An incoming session to SQL Server is classified, e.g., as a report server query.

The session is identified through its user-defined classifier function via Policy Management.
The session is routed to a workload group, e.g., Report_Server.
The workgroup will use the resource pool defined to it, e.g., Report_Server_Pool.
The resource pool enforces the limits required by the defined application.
An illustrative example involves preventing users in a group called MassiveQueryFun from using over
40% of the SQL Servers CPU time when running queries. The group in question is restricted much like
a restrictor plate on an engine in NASCAR racing. The restriction or policy element is the classifier that
activates the thresholds.
The resource governor is accessed via the Management Node of the SQL Server Management Studio.
Setup can occur via T-SQL statements. The elements of defining a resource with limits are the resource
pool, the workload group settings, and then the creation of classification functions for the resource pool.
Once the resource classification is registered with the resource governor then enable it and monitor the
performance. Here the DMVs allow quick query information about the resource use. The example below
establishes three workload groups under the default system resource pool and assigns the SQL Server
login power to the group, while establishing a classifier and then enabling it.
Figure 45: Resource Governor Setup T-SQL Statements

Exam Prep: Using the Resource Governor
You are responsible for managing an SQL Server 2008 server. A group exists called CrazyEddiesQueries.
This group conducts resource-intensive queries. You are a smart SQL administrator so you create a
workload group in Resource Governor and monitor the performance via a DMV query when members of
CrazyEddiesQueries execute a query.
The CPU use during this query period hovers around 92%. You determine that CrazyEddiesQueries must
be limited. What should you do?
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Exam Solution: Using the Resource Governor

In this case, most of the elements of limiting the CPU are in place. You have a workload group in place;
however, a classifier is not established to limit members of the CrazyEddiesQueries group from using too
much CPU time. You will create the classifier function to ensure the CPU limits are enforced.
You can quickly eliminate the exam answers that do not associate any workload groups with internal
resource pools. SQL Server as its own instance uses the internal resource pools, and this wont work for
user defined classifier functions. Rebuilding indexes always seems like a good solution to a performance
issue, but this could use MORE CPU time than it would reduce, so avoid any index building activities.
Max degree of parallelism settings also figure into possible solutions for this type of problem. The max
degree of parallelism settings applies to the number of server processors to use in executing a query
versus addressing individual queries. This is again a system wide setting that wont help address specific
queries that require boundaries.
In summary, always think of the resource governor in terms of its elements. Create a resource pool with a
given name. Assign the values for minimum and maximum CPU or memory use. Create a workload group
to use the resource pool. Then the established classifier function will return the name of the workload
group in which SQL Server will execute the query.
Database Engine Tuning Advisor

The database engine tuning tool is another weapon in the battle to optimize SQL Server. This is a
graphical-based tool that runs from SQL Server Management Studio and provides recommendations
based on settings selected for optimizing a database. The main focus of the database engine tuning tool
is recommending the best configuration with indexes, index views, and partitions. The tool conducts
processes that would be quite complex to analyze manually.
The tool analyzes collected database workloads from trace files or tables collected via the
SQL Server Profiler. The recommendations of the database engine tuning tool include:
QQ
Best configuration of indexes for databases based on query optimization
QQ
Suggests aligned or non-aligned partitions
QQ
Suggests index views
QQ
LearnSmart
Conducts analysis of the possible changes, in effect a query test for index use, spanning the
query across tables, and actual query performance
QQ
Allows input of design choices to affect the actual database design
QQ
Small query support
QQ
Provides a complete array of reports documenting recommendations
QQ
Allows some simulation of database server conditions, e.g., CPU constraints
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The basic steps for database engine tuning are:

1.
2.
3.
4.
5.
Create a workload for the database tuning activity either in T-SQL script by collecting a trace file
or collecting trace data to a table. This should represent the work conducted by the database
you want to tune.
Determine what you want to tune: indexes, index views, or partitioning best practices.
The database engine tuning output is in the form of logs, summaries, and recommendations
rolled up into reports.
If necessary, you can use the results to perform hypothetical modeling by changing parameters
and analyzing them.
Once you agree with tuning recommendations you can apply them to the database in question,
or schedule for a later time.
Figure 46: SQL Server Profiler Trace File Using Database Tuning Template
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 47: SQL Server Database Tuning Advisor Report Recommendations

Security is a consideration when using database tuning. Database tuning is activated by a member of the
sysadmin role. Once database tuning has been initialized by a member of the sysadmin role, then you can
extend the right to run database tuning to logins that have the db_owner role for a database. Any user
who is a member of the db_owner role can run the database tuning advisor for that database.
Test Tip: The database tuning advisor is also executable via the command line with dta.exe versus
using the graphical flavor of it. The command version of dta.exe has an extensive list of syntax switches,
which are easily accessed by typing dta.exe /?.
Exam Prep: Using Database Tuning Advisor
You manage an SQL Server 2008 instance called Racing\Performance. You create a workload with the
SQL Server Profiler and analyze it with the database tuning advisor. Two statements run by users Mufflers
and Sparks are not tuned. You need to tune these statements, while maintaining security. How will
you achieve this?
Exam Solution: Using Database Tuning Advisor
In this question we are asked to suppress login names within the database tuning advisor. The reason is
that when user names exist in the database tuning advisor it has impersonated the user when running
the statements in question. For database tuning to work it must be able to access the execution plan,
which is possible via the SHOWPLAN permission. The column named LoginName should be removed,
which will allow database tuning to make recommendations as it impersonates a user with the
proper credentials. In any type of security context question on the exam with database tuning advisor
avoid answers that grant too much security and those that grant security with no relation to the issue
at hand. An example is an answer that indicates you should assign users to the processadmin role to
facilitate database tuning recommendations.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Analyzing Trace Data with SQL Server Profiler

The SQL Server Profiler provides a great tool to capture the data needed to optimize SQL Server database
operations and ensure you can respond in a timely fashion to performance challenges. The exam
objectives expect the ability to configure SQL Server Profiler to gather data for specific problem types and
ensure you can filter the data and collect what is needed in relation to the problem under investigation.
The SQL Server profiler is accessed via the Performance Tools program group, or by clicking on Tools in the
SQL Server Management Studio.
The security required to run the SQL Server Profiler is the same as the T-SQL statements used to
create traces. Users must have at least the ALTER TRACE permission. One security note is that if a user has
VIEW SERVER STATE, ALTER TRACE, and SHOWPLAN permissions they can view the login and other security
information in the traces. This security aspect is important to note while evaluating exam questions. It is
very applicable when analyzing queries with SHOWPLAN results on.
SQL Server Profiler comes with several default templates for monitoring and includes ones for monitoring
stored procedures, T-SQL execution, and tuning. Any SQL Profiler monitoring session can be saved as a
template for future use. Saved SQL Profiler templates can be exported and imported by other SQL Server
Profiler instances. The SQL Server Profiler includes a deadlock graph option to help identify deadlocks
between processes. SQL Server Profiler traces can be saved to the Windows Server file system or to an SQL
Server table. SQL Server Profiler traces saved in tables can be queries for additional event data.
Figure 46 illustrates the SQL Server Profiler in action, with event class recording.
SQL Server Profiler is commonly used to:
QQ
Identify poorly written or performing queries
QQ
Find deadlock causes
QQ
Identify the performance characteristics of stored procedure execution
QQ
Monitor SQL Server audit activity, such as logins or logoffs
QQ
Show granular user based activity with T-SQL use
QQ
Collect data for performance testing or other replay activities
QQ
Collect data for use in database tuning activities
Exam Prep: Using SQL Server Profiler

You manage an SQL Server 2008 instance. An application team has asked for an XML formatted report that
shows processes and resources that are participating in deadlocks. How can you provide this quickly?
Exam Solution: Using SQL Server Profiler
The solution is to run an SQL Server Profiler session and select the Deadlock graph event under the
Locks events section. This supports collecting information about resources and processes that have
become deadlocked. The SQL Server Profiler supports extraction of the Deadlock events to a
separate XML file.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 48: SQL Server Profiler XML Extract Screen for Deadlock Events
Analyzing Performance Data

with Dynamic Management Views (DMV)
SQL Server DMVs provide a rapid way to find out the current state of an SQL operation. DMVs such as sys.
dm_exec_query_stats provide instant point in time data on the state of queries, while other DMVs show
wait states and many other elements. The use of the DMV is predicated on a snapshot type data need
versus sustained trending provided by tools like Performance Monitor and SQL Server Profiler.
Queries with DMVs will return server state information and can provide the data required for diagnosing
issues or tuning performance. DMVs are grouped into sections of use and span the entire range of SQL
Server functions.
DMV Categories
LearnSmart
QQ
Security
QQ
Transactions
QQ
Operating System
QQ
Extended Events
QQ
Service Broker
QQ
Resource Governor
QQ
Replication
QQ
Query Notifications
QQ
Objects
QQ
I/O
QQ
Indexes
Practice Exams
Audiobooks
Exam Manuals

n
QQ
Full-Text Search
QQ
Execution
QQ
Database
QQ
Mirroring
QQ
CLR
QQ
Change Data/Audit
1-800-418-6789
When using DMVs make sure you understand that there are thousands of possible T-SQL
combinations of DMV command syntax (this will be covered on the exam). However, always prefix the
name of the view or function by using the sys. schema. Then add the syntax for the function you query.
Security considerations require a minimum of SELECT permission on the object and VIEW SERVER STATE or
VIEW DATABASE STATE permissions.
A sample DMV query with results is shown below.
Figure 49: DMV of System Wait States
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Exam Prep: Using DMV for Optimizing SQL Server

You manage a server running SQL Server 2008 Enterprise. The accounting department has reported a
performance issue with their queries. They report they are running for 30 minutes versus the normal time
of 2 minutes. You have to track down the syntax of this long running query. How will you solve this?
Exam Solution: Using DMV for Optimizing SQL Server
The DMV sys.dm_exec_query_stats provides information about queries and lists their running times.
This will provide the name of the query that is running too long. Once you have the name identified with
the above DMV, then you can specify a SELECT statement to identify the syntax of the query. Use the
syntax SELECT * FROM sys.dm_exec_sql_text(<sql_handle>). This will identify the syntax of the query.
Once you have the syntax you can analyze why the query is running so long. This type of DMV question
on the exam expects an understanding of SQL Server T-SQL clause formatting. Some possible answers will
provide the proper syntax, but not have required actors present such as (). For example, this is a possible
solution answer: SELECT * FROM sys.dm_exec_sql_text<sql_handle>. However, examine this solution
closely and notice the clause is not valid. In this case you should avoid the use of WHERE clauses as this
is an argument not a search. Other common exam answers in this area provide a DMV in the answer that
does not correspond to the problem at hand. In this case a long running query will not require the use
of sys.dm_exec_query_optimizer_info, for example, as this DMV is related to database work tuning,
not process running queries. Any variation of the hundreds of DMV categories are fair game on the exam,
but you can optimize your test chances by ensuring answers follow the conventions of SQL T-SQL syntax,
and the DMVs apply to the issue in the question.
Performance Monitoring With the System Monitor

The System Monitor available with Windows 2008 Server is another tool you can use to troubleshoot
problems that are related to the server operating system and its hardware. The System Monitor has a
large amount of available counters or classes of attributes that provide an array of detailed information.
The installation of SQL Server 2008 on a Windows 2008 server automatically adds in a class of
performance counters for use in analyzing parameters associated specifically with SQL Server.
System Monitor is accessed via the Performance Monitor in the Administrative Tools program group.
The System Monitor allows a collection of data in a snapshot mode and the trending of data and
recording of it to database tables or file format. The use of System Monitor and its current incarnation in
Windows 2008 Performance Monitor is an entire subject and science that eclipses the scope of
this exam manual.
System monitor use is primarily related to trending memory use, CPU use, disk I/O, and the large amount
of counter objects available with each specific Windows 2008 Server subsystem or installed applications.
Interpreting the results of System Monitoring data is the goal of the questions in this area of the exam.
If you are monitoring server RAM and Process: Working Set of RAM shows Maximum use at 4 GB and your
server only has 4 GB of RAM total then you know a problem exists. You must also correlate settings from
SQL Server to what is observed in System Monitoring collected data. You become like a detective in that
you must be able to pinpoint what performance issues are occuring and how they relate to SQL Server.
Keep in mind that on the exam the primary hardware influences on SQL Server devolve to RAM,
Disk speed, and CPU use.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals
1-800-418-6789
Exam Prep: Using Performance Monitor to Optimize SQL Server

Your SQL Server 2008 server has 16 GB of RAM and supports several other applications.
SQL Server Memory Configuration
Min Server Memory = 4 GB
Max Server Memory =8 GB
The Sales group complains often of poor performance when they use other applications on the server.
How can you improve performance without any hardware investments?
System Monitor Counters:
Counter Object
Minimum
Maximum
Process: Working Set
2 GB
3 GB
SQL Server: Buffer Manager: Buffer

Cache Hit Ratio
90
97
SQL Server: Memory Manager: Total

Server Memory (KB)
12 GB
16 GB
Exam Solution: Using Performance Monitor to Optimize SQL Server

Like so many exam questions always make sure you read the question closely. In this case the question
does not ask you to do anything to increase SQL Server performance. This question poses a situation
where SQL Server is performing fine, but causing other server based applications to suffer. The counters
reveal that SQL Server is only using a maximum of 3 GB of memory, yet configured to take a minimum of
4 GB from the operating systems pool of memory. A simple decrease of the SQL Min Server Memory
to 3 GB will allocate 1 GB back to the servers operating memory pool and make it available to other
server applications. This type of question specifies that no hardware investment occur in the solution.
One answer will invariably propose adding RAM, which is immediately ruled out. Any solutions that
propose increasing resources to SQL Server are also wrong.
Domain 8 Implementing High Availability

This collection of software and hardware based solutions for high availability combine to provide SQL
Server the capability for continuous operations, without downtime. This exam domain expects fluency
with each method, what situations to use them in, and how to implement and manage them.
SQL Server provides several methods to ensure continuous operations. Database mirroring is an inherent
SQL Server 2008 solution, and it allows the replication of all transactions to a second database with
monitoring to decide when it should take over. SQL Server is completely API aware of Microsoft Windows
Server 2008 R2s cluster capabilities, and it will install as a native service. SQL Server clustering using the
features of Windows 2008 R2 server provides the ability to protect against single server hardware failure,
disk failure, or a failure of the SQL Server software bits. Transaction log shipping provides protection at
the SQL Server layer by supporting database movement of all transactions to a second database, which
is in a single use mode and ready to become the primary in case of failure. SQL replication supports the
synchronization of data between databases in a consistent state. Data is distributed via replication to
different SQL Server databases, and when necessary, to mobile clients.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Database Mirroring
SQL 2008 database mirroring is a database availability solution that literally mirrors all data from one
server database to another, provided each database is on a different SQL Server instance. One server is
a primary copy or one in use, and the other is termed a standby server. The SQL Server 2008 database
mirroring software and administration functions provide for the synchronization of data between the
two databases in such a way that no transaction is every dropped during the failover to the secondary
standby database. The database mirror can also use a standby database copy without up to date
synchronization, which can result in some data loss, but still provide a server ready to pick up
database operations in the event of a failure to the primary.
Figure 50: Database Mirroring Schematic
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
The figure above demonstrates the elements of a database mirroring configuration:

QQ
QQ
QQ
QQ
Two SQL instances are identified, and then a database is selected for mirroring (in this case
the db Parts).
The mirroring wizard is run via SQL Server Management Studio, or you can use ALTER DB to
setup mirroring via T-SQL.
A mode is selected for the mirroring operation. In most cases, for complete recoverability, use the
High Safety mode with automatic failover, which requires a witness server.
The witness function works much like the witness server in Windows 2008 Server clustering and
the quorum drive concept:

If the witness server loses connection to the primary mirror server (Headers01) and is
connected to Headers02, a failover will occur (i.e., the second database server in the
mirroring configuration will serve as the primary database instance).

When the server Headers01 comes back online it will receive a message from the
witness server that Headers02 is now the primary server in the mirror setup, and this
will cause Headers01 to become the mirror database and to synchronize transactions.
This is a very elegant solution and one that in production situations induces 2-5 seconds of connectivity
for applications such as .NET web applications.
Exam Prep: SQL Server Database Mirroring
You manage multiple SQL Server 2008 servers for your company. You set up database mirroring. You use
automatic failover. An unexplained error occurs in the primary database in the mirrored databases.
The current database state is synchronizing. The witness state is not working. You need to act quickly to
fail the mirror over, and data loss is acceptable.
Exam Solution: SQL Server Database Mirroring
This exam style question immediately should prompt you to know the method of failover is now the
forced method. The forced method is one used to force failover, but does not guarantee that data loss will
not occur. This method of forcing a service change is normally used in a disaster recovery situation,
where the primary member of the mirror may never come back online. One other clue in the question
pointing to the direness of the situation is that the current database state is synchronizing, which means
they are not synchronized anymore.
Manual failovers require both the primary and mirror server to be online and the database to
be synchronized. Replacing the mirror server is not a good solution at any point in SQL Server database
mirroring operations.
There are many security considerations with SQL Server database mirroring. One upper level concern is if
users are going to login to SQL Server databases with SQL credentials, then you must ensure you create
all logins and database mappings in the mirror target SQL Server because you cannot mirror the master
database on the primary mirror server, which houses all logins and mappings. Database object users and
assignments will exist due to mirroring.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
SQL Server Clusters

Microsoft SQL Server is a resource for clustering with Microsoft Windows 2008 R2 clustering services.
This type of cluster provides complete failover in protecting the SQL Server instance from a failure due
to any type of hardware failure. In a Microsoft Windows 2008 cluster implementation, the SQL Server
instance uses a shared disk resource normally located on a SAN and is set up from the start as a cluster
from the SQL Server installation. A SQL Server cluster consists of at least two servers configured to run
an SQL instance that is configured to work on both servers. The SQL cluster implementation creates a
virtual SQL Server with its own IP address and dedicated set of disk resources. The SQL Server clusteraware installation will load the SQL bits to all servers identified in the cluster setup at the same time
during installation. Windows quorum uses a quorum and witness concept to determine which cluster
members are viable and can participate in supporting cluster resources. Windows clustering uses a
heartbeat based monitoring system to constantly interrogate the status of cluster members. In best
practice cases the Windows cluster node heartbeat network interface cards are dedicated to heartbeat
operations and set up on a private VLAN for traffic segregation to ensure their function is not encumbered
by normal Windows cluster server operations. The emergence of Windows 2008 Hyper-V virtualization
services has added a new dimension to cluster operations with the capability of managing cluster
members as virtual guests. Windows 2008 Hyper-V hosts can also be clustered at this time with Windows
2008 cluster services.
Test Tip: A SQL Server cluster using Windows 2008 R2 clustering features is managed from the Microsoft
Windows 2008 R2 cluster administrator utility, NOT from SQL Server Configuration for tasks such as
restarting the SQL Service.
Figure 51: SQL Server Cluster Implementation
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 51 shows a simple Windows 2008 R2 SQL cluster with active/passive configuration. The Windows
cluster is comprised of two nodes: Cluster01 and Cluster02. Once Windows 2008 R2 clustering is
configured, a third virtual or cluster node is created, for example, ClusterNode. It is, however, comprised of
the two physical servers listed above. This cluster can now provide failover for standard Windows services:
DHCP, file sharing, disk drives, and other cluster available resources.
Once a cluster is established and functional, a SQL Server installation is possible that will detect
the Windows cluster and allow the running of the setup path to establish a SQL cluster instance.
This will create the virtual SQL cluster SQL01 with its own IP address. The SQL cluster instance will
have several prerequisite resources to launch, the primary one being the disk drives that are cluster
accessible and meet the definition of a shared cluster resource. The virtual SQL cluster instance named
SQL01 is what clients and applications would target in their connection string definitions and for
other database operations.
The failover process can identify which cluster nodes are available as possible owners of the cluster
SQL instance. If Cluster02 was a failover host for Cluster01, and Cluster01 suffered a bluescreen/memory
dump issue, then a failover of all SQL instances and services would occur to Cluster02.
This is amazing technology and represents a top-level solution for protecting SQL Server availability.
The failover process also provides quite a bit of management support; you can update and patch
servers in a cluster, as required, without downtime as you failover resources to allow for the updates
to occur. Server reboots and other types of maintenance such as a RAM addition can also occur during
the production cycle of a day due to failover capability.
Figure 52: Cluster Resources for an SQL Server Cluster

Figure 52 shows a SQL Server cluster implementation with the established SQL Server cluster group and
member cluster resources such as Analysis Service, DTC, SAN Drives, SQL Server virtual IP, SQL Server
virtual name, the SQL Server service, SQL Server Agent, and full-text search. These cluster resources
combine to form the SQL cluster instance.
Again, if the S Drive above is not online, then the SQL cluster will not start as it is a dependent resource
for the SQL Server service. For security, each node in an SQL cluster will use the same service accounts to
run the SQL Server service. This account should be protected and managed properly to ensure it does not
lockout or become unavailable, as this will cause the SQL cluster to fail.
Exam Prep: SQL Server Clustering
You support six instances of SQL Server 2008 running on different computers configured as
failover clusters. You are required to add a seventh SQL node to the Windows 2008 failover cluster.
The least amount of effort should be used. How will you accomplish this?
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Exam Solution: SQL Server Clustering

This type of question will present several solutions that are designed to test your knowledge of how to
initiate a clustered SQL instance. To configure an SQL Server instance as a node in a cluster you must run
the SQL Server 2008 setup. This will present you with the option of adding a node to a cluster. This is the
same for removing a node from a cluster. The SQL Server Management Studio is of no use here,
and neither is a T-SQL based solution like sp_configure.
Transaction Log Shipping

Transaction log shipping is an older SQL Server method of high availability but one that does not have the
expense of SQL Server clustering. It does not have the rapid failover capability of SQL database mirroring
but is very viable to ensure your SQL database transactions are replicated and available for use in the
event of a problem to the database setup as the log shipping primary.
Log shipping automatically sends transactions from a primary database to a secondary database,
on separate SQL Server instances. The log shipped data is continuously applied to the secondary
database as it arrives. An option exists to use a third SQL instance to monitor the history and status
of all log shipping operations between the primary and secondary database.
Test Tip: Log shipping secondary databases do not automatically come online for a failover operation
if the primary database fails. You must start the secondary database with a full restore of the primary
database backup using NORECOVERY or STANDBY.
Log shipping conducts a series of SQL jobs to achieve the movement of transactions. A backup job
occurs on the primary database every 15 minutes, but this is configurable. A copy job copies the primary
database backups to the secondary server. The job is called Log Shipping Copy and is created when log
shipping is created on the secondary server. A restore job is also created on the secondary server,
which contains the copied backup files, and all the relevant backup history.
During a log shipping recovery operation the following tasks are conducted:
QQ
Apply any uncommitted backups from the backup file share established for log shipping.
QQ
Apply any uncommitted transaction log backups to the secondary database.
QQ
Conduct a tail log backup of the primary database if it is available.
QQ
Recover the secondary database to place it in a consistent state, and bring it online.
QQ
QQ
LearnSmart
Advise clients of the database instance name where the previously primary database
is now running.
If desired, set up other secondary log shipping targets for the new primary database.
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Figure 53: Log Shipping Schematic

Exam Prep: SQL Server Log Shipping
You manage several SQL Server 2008 servers. Headers02 is configured as a secondary log shipping server
for Headers01. You are conducting maintenance on Headers01 and need to failover first to Headers02.
What step should you perform first?
Exam Solution: SQL Server Log Shipping
In this case you would back up the transaction log on Headers01 with the NORECOVERY option, which will
get the tail of the transaction log and allow you to bring the log shipping secondary database up to date
with transactions it may not have from the previous logs shipped. This is planned cutover, so there is not
a physical problem with the primary server, which would dictate different actions on the log shipping
target server. Some exam solutions appear valid but ensure the above options which direct you
to conduct a backup of Headers01 with the COPY ONLY option as this will not provide the tail log
information needed by the secondary.
SQL Replication
SQL Replication provides SQL Server the capability of pushing data to multiple endpoints in a
controlled and synchronized fashion. SQL Replication can replicate data between SQL Server and a
client, such as a remote inventory system or mobile notebook users. In the high availability space, SQL
Replication can replicate data from SQL instance to SQL instance. This can also include supporting data
warehousing or reporting.
SQL Server replication uses a push or pull type model to replicate data predicated on standard
taxonomies of publisher, subscriber, publications, articles and subscriptions. The publisher is a source
database pushing data out to other instances. The distributor is a database store that houses transaction
data about the replication process to include replication status, metadata, and in some cases queues
data for subscribers. The publisher and distributor can be the same database. Subscribers are databases
that receive the publishers data. A subscriber can be bidirectional in that it passes data back and forth
to publishers. An article is a database object in a publication. This can be views, tables, or indexes.
Data published as articles supports SQL Replication filtering, which is important for determining the
data composition of what is moved. A publication is a grouping or article, which can be from more than
one database. Lastly, the subscription is a request for data from a subscriber. The subscription determines
what will be received, when, and where. Subscriptions are push or pull type replication.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
SQL Replication is a feature that supports the following types of publication:

QQ
QQ
QQ
Merge Publication - A snapshot of the data occurs and then changes are tracked
with triggers. When subscribers connect up for the changes, all changes are published
since the previous replication.
Transactional Publication/Transactional Publication With Updatable Subscriptions -
A snapshot of the data occurs at the publication database and then any changes are delivered to
the subscribers as they occur. With updatable subscriptions, changes made at the subscriber are
applied back to the publisher.
Snapshot Replication Snapshots are point in time replications and no updates are tracked.
Figure 54: SQL Server Replication Using Multiple Methods of Replication
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Exam Prep: SQL Server Replication

You manage multiple SQL Server 2008 instances. You have a main office in Atlanta and a branch office
in Seattle. The branch office is connected to corporate via a WAN connection. There is one SQL Server
2008 instance per office. The SQL Server 2008 instance in Atlanta contains a database named Parts,
and a table with 2 million rows named Parts_Stock. You set up merge replication between the Parts_Stock
table and a table named Parts_Stock in the remote database instance located in Seattle. A WAN failure for
two days causes the most recent replication merge to expire. You need to get the remote copy of Parts_
Stock up to date as quickly as possible. You should complete this without much effort. How will you
solve this problem?
Exam Solution: SQL Server Replication
In this case you will create a new subscription. A tablediff will identify the table changes and use the
new subscription to apply the differences. You will not need to create a snapshot. This is the least cost in
terms of bandwidth and effort for a solution. Exam answers that focus on the publication portion of the
replication setup in this case expend too much effort because it is harder to set up the publication side
than a subscription.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Practice Questions
Chapter 1
1.

You work as the DBA for a medium-sized organization. One of the employees in the Accounting
department has asked for your assistance. She needs a solution that will allow her to transfer
data from a MySQL database into a SQL Server database. Additionally, the data must be modified
in several ways during the process. Which SQL Server add-on component will meet her needs?
Select the best answer.
A. SSAS
B. SSRS
C. SSIS
D. Database Mirroring
2.

You are configuring database mail and have been asked to ensure that fault tolerance is
provided in order to accommodate for random mail server outages. Which of the following
actions can provide this fault tolerance? Select the best answer.
A. Set the logging level to verbose
B. Configure multiple mail profiles
C. Configure the account retry delay to 90 seconds
D. Set the database executable minimum lifetime to 900 seconds
3.

You are using a mail server to send mail from SQL Server 2008s database mail feature. The mail
server has intermittent problems that may cause it to be unavailable for 30 to 90 seconds.
You cannot use any other mail servers. What database mail configuration parameters may
help ensure that email messages are delivered? Choose TWO.
A. Account retry delay
B. Create additional profiles
C. Logging level
D. Account retry attempts
Chapter 2
1.

When you configure an alert in the SQL Server Agent, only two response options are provided.
What are these TWO options? Choose TWO.
A. Execute job
B. Notify operator
C. Shutdown server
D. Check the integrity of a database
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
2.

1-800-418-6789
You are configuring an alert. Based on the screen in Exhibit 1, what is the value of buffer cache
hit ratio that would fire the alert? Select the best answer.
A. Less than 7.8 percent
B. Less than 78 percent
C. Greater than 78 percent
D. Greater than 7.8 percent
Exhibit(s):
3.

You manage a single SQL Server 2008 instance with a single database called Engineering that
is in the Simple recovery mode. You have been creating Full backups every night for
several months. Now, the database has grown so large that the full backups take more than
two hours. You want to perform Full backups every weekend and a different type of
backup nightly. You must use the built-in backup feature of SQL Server 2008. What kind of
backup can you perform that cooperates with the weekend Full backups, but requires less time?
A. Incremental
B. Modified
C. Transaction Logs
D. Differential
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Chapter 3
1.

You are the DBA for a large SQL Server 2008 database application. When you arrived at work this
morning, you had several messages waiting from users. The users were complaining that they
cannot access the data. They are receiving access denied errors. All users could access the data
the previous day. At the end of the day yesterday, you performed the following tasks:-Manual
backup of the database-Manual snapshot of the database-Modified the ownership of the
call_routing table-Added seven new usersWhich action is the likely cause of the problem?
A. Manual backup of the database
B. Manual snapshot of the database
C. Added seven new users
D. Modified the ownership of the call_routing table
2.

You are running SQL Server 2008 Enterprise Edition. You want to enable a server-level audit
procedure using SQL Server Audit. What THREE steps are required to implement a server-level
audit procedure? Choose THREE.
A. Create a server audit specification
B. Enable the audit
C. Create an audit object
D. Create the server-level permission for the audit
3.

You are installing a test server that will run SQL Server 2008 Enterprise Edition. The installation
OS will be Windows Server 2008. You want to ensure that remote connections are not allowed.
How will you configure the installation so that only local connections are allowed but outgoing
network connections can be made? Select the best answer.
A. Unplug the network cable
B. Disable the TCP/IP provider
C. Block TCP port 21 in the firewall
D. Use the sp_configure Block Net-Conn, 1 command
Chapter 4
1.

You are creating a backup plan. Part of the plan includes the backup of the transaction logs for
your databases. All databases run on SQL Server 2008 Standard Edition and the OS used is
Windows Server 2008. You plan to use the following code to backup the transaction logs:
BACKUP LOG database_name TO DISK=logback.bak;In addition to backing up the log, what
would this code do? Select the best answer.
A. Truncate the log
B. Backup the database
C. Restore the database
D. Truncate the database
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
2.

1-800-418-6789
You are the DBA for a small company in Florida. You must restore a database named Sales to 3:17
yesterday afternoon. The current date is Wednesday, June 3, 2009. A full backup is made every
night to a file named day-fback.bak. The transaction log is backed up each day at 12 PM and is
backed up to a file named day-log.bak. Each week, the files are overwritten. Which THREE of the
following four commands would accurately restore the database to the desired point-in-time
if run in the proper order? Choose THREE answers.
A. RESTORE LOG Sales FROM DISK=TUESDAY-log.bak WITH NORECOVERY, STOPAT = Jun 2,

2009, 3:17 PM;
B. RESTORE DATABASE Sales WITH RECOVERY;
C. RESTORE DATABASE Sales FROM DISK=TUESDAY-fback.bak WITH NORECOVERY;
D. RESTORE LOG Sales FROM DISK=TUESDAY-log.bak WITH NORECOVERY, UNTIL = Jun 2,

2009, 3:17 PM;
3.

You are creating a Maintenance Plan using the wizard. You want to make sure the database is not
corrupt and perform a complete backup of the database. In Exhibit 1, what options should
you check? Choose TWO.
A. Rebuild Index
B. Back Up Database (Full)
C. Shrink Database
D. Check Database Integrity
Exhibit(s):
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Chapter 5
1.

You are the DBA for a carefully planned database. In spite of the intensive planning, the data
files are still growing very large. Youve been asked to look for a solution that reduces the storage
space consumed by the database. The performance of the database is not an issue, as the users
have been very happy with the performance; however, space is quickly running out on the
server and the server administrator has indicated that more drive space cannot be purchased
until the next budget cycle. What feature can be used to conserve storage space?
A. Data compression
B. Transparent data encryption
C. Table partitioning
D. Filestreams
2.
What kinds of pages exist in a SQL Server index? Choose THREE.

A. Intermediate pages
B. Leaf pages
C. Root pages
D. Branch pages
3.

You are attempting to create an index with the WITH ONLINE = ON clause. However, it is
not working. You are running the SQL Server instance on Windows Server 2008 Standard Edition
and the SQL Server instance is running the Standard Edition of SQL Server 2008. Service pack
1 has not been applied to the Windows Server. Several users are connected to the database.
Why is the index command not working? Select the best answer.
A. Because Windows Server 2008 Enterprise Edition is required
B. Because users are connected
C. Because service pack 1 has not been applied
D. Because SQL Server 2008 Enterprise Edition is required
4.

A database uses the collation named French_CI_AI. Without looking at documentation,

what would this collation name tell you by itself? Choose three.
A. Case insensitivity
B. Accent insensitivity
C. The French codepage is used
D. Albanian icons
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Chapter 6
1.

You must configure a SQL Server Agent job to use a proxy account for access to the Windows
file system. The job runs on a SQL Server 2008 Standard Edition server. You have taken the
following steps:1. Right-click on the Credentials folder and select New Credential.2. In the New
Credential windows enter a Credential name of OSCommand and select a user with appropriate
permissions as the identity and provide a password.3. Click OK to create the Credential.When
you attempt to create the job step and use the proxy account, you see the Run As option
shown in Exhibit 1. What steps do you need to take to use the proxy account based on the
OSCommand credential? Select the best answer.
A. You must use the CREATE JOB STEP command since proxies are not available in

the GUI interface.
B. Proxies are only supported on Enterprise Edition.
C. Create a new Operating System (CmdExec) proxy in the Proxies container within the SQL

Server Agent node.
D. The proxy account is configured as the owner of the job and not using the Run As

parameter of a step.
Exhibit(s):
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
2.

1-800-418-6789
You are creating a proxy object for use with a SQL Server Agent job. You want a single object to
provide all needed access for the job. The job runs several steps including the following
actions:-Powershell-Batch files-VBScript scripts-Transact-SQLWhat SQL Server Agent subsystems
should be enabled for the proxy object? Choose all that apply.
A. SQL
B. Operating System (CmdExec)
C. ActiveX Script
D. PowerShell
3.

Which of the following tools can be used to see the most recent run of a SQL Server Agent job?
Choose all that apply.
A. Configuration Manager
B. Job Activity Monitor
C. Job Manager
D. Log File Viewer
Chapter 7
1.

Which of the following are valid recommendations that DTA can make for performance
improvements after a workload analysis? Choose all that apply.
A. Adding indexes
B. Reserved memory recommendations
C. Deleting indexes
D. Partitioning tables
2.

You want to schedule a Database Engine Tuning Advisor analysis to run during the night.
You plan to implement a SQL Server Agent job. On what command line tool must the job call to
perform the analysis? Select the best answer.
A. DETA.EXE
B. DTA.VBS
C. DETA.VBS
D. DTA.EXE
3.

Which one of the following is NOT an item that you define within a SQL Server Profiler trace?
A. Events
B. Stored Procedure
C. Data columns
D. Filters
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
4.

1-800-418-6789
You want to evaluate fragmentation statistics for each index in a database. Which DMV will
you query? Select the best answer.
A. sys.dm_db_index_operational_stats
B. sys.dm_db_index_fragments
C. sys.dm_db_index_physical_stats
D. sys.dm_db_index_physical_fragments
Chapter 8
1.

You are planning a database mirroring implementation. Which roles, related to mirroring, must
be configured on separate SQL Server instances? Choose all that apply.
A. Primary
B. Witness
C. Mirror
D. Reflection
2.

You are implementing replication. High availability is important. Youve chosen to implement
replication with the following options:-Immediate Updating Subscribers-Publisher-Distributor-
Subscriber Which one should be changed or removed to allow for high availability?
A. Publisher
B. Distributor
C. Immediate Updating Subscribers
D. Subscriber
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Answers & Explanations

Chapter 1
1. Answer: C
Explanation A. Incorrect. SQL Server Analysis Services (SSAS) is used to analyze data and perform business
intelligence operations.
Explanation B. Incorrect. SQL Server Reporting Services (SSRS) is used to generate reports against
SQL Server data.
Explanation C. Correct. SQL Server Integration Services (SSIS) will provide extraction,
transformation and loading.
Explanation D. Incorrect. Database mirroring is a high availability solution that creates an exact duplicate
(mirror) of a database on a separate server.
2. Answer: B
Explanation A. Incorrect. Configuring the logging level to verbose will provide more information about a
failure, but it does not provide for fault tolerance in any way.
Explanation B. Correct. Database mail allows for the configuration of multiple mail profiles so that a
secondary or tertiary profile may be used if the first mail profile is not functioning.
Explanation C. Incorrect. Setting the account retry delay to 90 seconds will potentially give the mail server
more time to become available again, but it does not create a fault tolerant scenario.
Explanation D. Incorrect. This setting determines how long the database mail send utility will remain
in memory. Keeping it in memory longer prevents reloading it again and again, but it does not improve
fault tolerance - even though it may improve performance.
3. Answers: A, D
Explanation A. Correct. The default for this setting is 60 seconds. In this scenario, setting the value to 90
seconds would increase the likelihood that the email could be sent on the second attempt.
Explanation B. Incorrect. Creating additional profiles will not help since you can only use the
one mail server.
Explanation C. Incorrect. Configuring the logging level only determines the detail of the logs.
Explanation D. Correct. By default the account will be retried only once. By increasing this value, you
ensure that the email will get out when the mail server becomes available.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Chapter 2
1. Answers: A, B
Explanation A. Correct. You can execute any pre-configured job when an alert fires.
Explanation B. Correct. You may notify an operator when an alert fires.
Explanation C. Incorrect. You may be able to shutdown the server through a job, but the alert can only call
on the job.
Explanation D. Incorrect. You can check database integrity through a job, but the alert can only call
on the job.
2. Answer: B
Explanation A. Incorrect. The correct response is less than 78 percent.
Explanation B. Correct. Anything less than 78 percent would fire this alert.
Explanation C. Incorrect. The correct response is less than 78 percent.
Explanation D. Incorrect. The correct response is less than 78 percent.
3. Answer: D
Explanation A. Incorrect. Incremental backups are not supported by the internal SQL Server backup tools.
Explanation B. Incorrect. No such backup feature exists in SQL Server.
Explanation C. Incorrect. Backing up the transaction logs for a database in the simple recovery mode
would not provide full recoverability.
Explanation D. Correct. Differential backups are used to backup only changed data pages in the database.
Chapter 3
1. Answer: D
Explanation A. Incorrect. Backing up the database should not impact permissions.
Explanation B. Incorrect. Creating a snapshot should not impact permissions.
Explanation C. Incorrect. Adding users should not cause existing users to lose access.
Explanation D. Correct. The change in ownership has likely resulted in a break in the ownership change.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
2. Answers: A, B, C
Explanation A. Correct. The server audit specification will specify the item to audit and map it to
the audit object.
Explanation B. Correct. The final step is to enable the audit.
Explanation C. Correct. The audit object will define the target of the audit.
Explanation D. Incorrect. Permissions are not required since the audit procedure runs within the SQL
Server database engine service.
3. Answer: B
Explanation A. Incorrect. This action would prevent outgoing network connections as well.
Explanation B. Correct. The TCP/IP provider enables remote connections. You can disable them by turning
off the provider in the SQL Server Configuration Manager.
Explanation C. Incorrect. While this might prevent FTP traffic, it will do nothing to prevent SQL Server
access on port 1433.
Explanation D. Incorrect. No such command exists.
Chapter 4
1. Answer: A
Explanation A. Correct. By default, without the use of the TRUNCATE_ONLY parameter, the log file is
truncated after the backup completes. This truncation means that the backed up portion of the log is
emptied for reuse.
Explanation B. Incorrect. The BACKUP LOG command never backs up the database.
Explanation C. Incorrect. The RESTORE DATABASE command is used to restore a database.
Explanation D. Incorrect. While the command will truncate the log file, it will not do so to the database.
2. Answers: A, B, C
Explanation A. Correct. This is the second step in the process.
Explanation B. Correct. This is the third step in the process.
Explanation C. Correct. This is the first step in the process.
Explanation D. Incorrect. The proper WITH parameter is STOPAT and not UNTIL.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
3. Answers: B, D
Explanation A. Incorrect. Rebuild Index will drop the indexes and regenerate them.
Explanation B. Correct. Back UP Database (Full) will perform a complete backup of the database.
Explanation C. Incorrect. Shrink Database will shrink the physical size of the database file based on free
space within the file.
Explanation D. Correct. Check Database Integrity will perform a DBCC CECKDB operation
against selected databases.
Chapter 5
1. Answer: A
Explanation A. Correct. As long as the server is running Enterprise Edition, data compression can be used.
Data compression is not available in Standard Edition.
Explanation B. Incorrect. TDE does not conserve storage space.
Explanation C. Incorrect. Partitioning does not conserve storage space.
Explanation D. Incorrect. Filestreams do not conserve storage space.
2. Answers: A, B, C
Explanation A. Correct. The intermediate pages contain branches leading to other intermediate pages
or leaf pages.
Explanation B. Correct. The leaf pages contain the actual data stored in the index.
Explanation C. Correct. Each index contains one and only one root page.
Explanation D. Incorrect. While a B-tree structure is used, the pages are not called branch pages.
3. Answer: D
Explanation A. Incorrect. SQL Server 2008 Enterprise Edition is required, but Windows Server 2008
Standard Edition can run the Enterprise Edition of SQL Server.
Explanation B. Incorrect. If SQL Server 2008 Enterprise Edition were in use, the users could be connected
and the command could run.
Explanation C. Incorrect. SP1 is not needed for online operations; however the Enterprise Edition of SQL
Server 2008 is required for online operations.
Explanation D. Correct. Online index operations require SQL Server 2008 Enterprise Edition.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
4. Answers: A, B, C
Explanation A. Correct. The CI stands for case insensitivity.
Explanation B. Correct. The AI stands for accent insensitivity.
Explanation C. Correct. The word French clearly denotes the French language.
Explanation D. Incorrect. AI does not stand for Albanian icons, it stands for accent insensitivity.
Chapter 6
1. Answer: C
Explanation A. Incorrect. A proxy object must be created that is based on the credentials configured in
these three steps.
Explanation B. Incorrect. Proxies are supported on both Enterprise and Standard editions.
Explanation C. Correct. Proxies are based on credentials. Creating the credential is only the first step.
Next, you need to expand the SQL Server Agent node and then the Proxies node. Now, right-click on
the Operating System (CmdExec) node and selectn New Proxy. Enter the proxy name, credential name
of OSCommand and check Operating System (CmdExec) in the Subsystem section. Click OK to
create the proxy.
Explanation D. Incorrect. The proxy account is configured using the Run As parameter; however,
a credential is the foundation for a proxy object. The proxy object must be created in the Proxies node of
the SQL Server Agent node.
2. Answers: B, C, D
Explanation A. Incorrect. Transact-SQL commands are handled internally without the need for a proxy.
Explanation B. Correct. The batch files will require CmdExec to operate.
Explanation C. Correct. The VBScript code will require the ActiveX Script subsystem to function.
Explanation D. Correct. The PowerShell subsystem is used to execute PowerShell commands.
3. Answers: B, D
Explanation A. Incorrect. The SSCM is used to configure SQL Server services and it does not show
information about job histories.
Explanation B. Correct. The Job Activity Monitor can be used to see that last run and whether it was
successful or not.
Explanation C. Incorrect. No such utility exists in SQL Server.
Explanation D. Correct. The Log File Viewer is displayed when you right-click a job and select View History.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Chapter 7
1. Answers: A, C, D
Explanation A. Correct. One of the simplest ways to improve database performance is to add
appropriate queries. DTA can recommend additional indexes.
Explanation B. Incorrect. DTA makes no recommendations related to instance configuration.
Explanation C. Correct. You can often improve performance by deleting indexes that are not
being utilized. DTA can recommend dropping indexes.
Explanation D. Correct. Table partitioning can be recommended by DTA, though it is not by default.
2. Answer: D
Explanation A. Incorrect. The acronym for the Database Engine Tuning Advisor is DTA and not DETA.
Therefore the command is DTA.EXE.
Explanation B. Incorrect. The command is an executable (DTA.EXE) and not a script.
Explanation C. Incorrect. The command is an executable (DTA.EXE) and not a script.
Explanation D. Correct. With DTA.EXE you can configure an analysis process within a job.
3. Answer: B
Explanation A. Incorrect. You select events when building the trace.
Explanation B. Correct. You can capture information about stored procedures, but they are not part of the
trace configuration.
Explanation C. Incorrect. You can select specific data columns related to the events when
building the trace.
Explanation D. Incorrect. You can create filters when building the trace.
4. Answer: C
Explanation A. Incorrect. This DMV reports locking, latching and access stats for the indexes
and not fragmentation.
Explanation B. Incorrect. No such DMV exists.
Explanation C. Correct. This DMV will report fragmentation statistics.
Explanation D. Incorrect. No such DMV exists.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

n
1-800-418-6789
Chapter 8
1. Answers: A, B, C
Explanation A. Correct. The Primary role must exist in a separate instance from the Mirror or Witness roles.
Explanation B. Correct. The Witness role must exist in a separate instance from the Primary
and Mirror roles.
Explanation C. Correct. The Mirror role must exist in a separate instance from the Primary
and Witness roles.
Explanation D. Incorrect. No such role exists in a mirror relationship.
2. Answer: C
Explanation A. Incorrect. The publisher is required. Without it, you will have no replication at all.
Explanation B. Incorrect. The distributor is required. It may be the same server as the publisher,
but it is required.
Explanation C. Correct. If high availability is important, you should use queued updating subscribers,
which can process transactions even if the publisher is unavailable.
Explanation D. Incorrect. If you have no subscribers, you have no need for replication.
LearnSmart
Practice Exams
Audiobooks
Exam Manuals

70-432

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

70-432

Uploaded by

Copyright:

Available Formats

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

SQL Server 2008 Implementation

Warning and Disclaimer

Volume, Corporate, and Educational Sales

International Contact Information

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Finding Error Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Domain 1 Installing and

Hardware and Software Requirements

32-bit, a 32-bit processor

64-bit, a 64-bit processor

IA64, a 64-bit custom

Microsoft SQL Versions

Microsoft SQL Server 2008

Microsoft SQL Server 2008

Microsoft SQL Server 2008

Microsoft SQL Server 2008

Microsoft SQL Server 2008

Table 1: SQL Server and Supported Architectures

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Required Disk Space

All versions of SQL Server 2008

32-bit versions of SQL Server

SQL Server has different disk size

One exception is that SQL

64-bit versions of SQL Server

Other SQL Server options will

Table 2: SQL Server Hardware Requirements

Installing SQL Server 2008

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Figure 1: SQL Server Install Options

SQL Server Instances

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Figure 2: Practical Examples of SQL Server Instances in Use

Selecting SQL Server Features

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Figure 3: SQL Server features during install

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

SQL Server Security Considerations During Install

Figure 4: SQL Server Installation Dialogue for Selection of Service Accounts

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Figure 5: SQL Server Account Provisioning and Authentication Mode Choices

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)

Figure 6: T-SQL Statement to complete FILESTREAM post-installation configuration

Data Directories in SQL Server Installation

Cloud Classroom: Video Training

Exam Manual SQL Server 2008 Implementation and Maintenance (70-432)